International Persuasive Communication: Concepts and Theory

What explains the ability or failure of a particular state to win support and approval among foreign populations? Nye would frame that question as what makes a state “attractive” and his answer would point to the given state’s policies, culture, and values, along with their effective communication. The “(re)sources” of Nye’s soft power are not all controlled by governments: he allows that civil society, non-governmental organisations and private commercial actors can all be involved in “generating attraction”’ for their country of origin. ^{Footnote 16} But government officials across the world are operating in the belief that their own state-funded soft-power projects can make a positive difference—and it is these efforts that have attracted most attention in the soft-power literature. A wealth of studies describe the measures state officials are taking in their desire to shape international public opinion, ^{Footnote 17} and political discourse surrounding these measures has also been researched. ^{Footnote 18} What is missing from this literature, however, is evidence-based theory regarding how perceptions of foreign countries take shape among mass publics. The exercise of soft power is discussed as ‘a long-term process that should be barely noticeable,” ^{Footnote 19} which almost excludes the prospect of tracing it empirically. Consequently, the soft-power concept has become in practice “a complex of assumptions about the modalities of influence,” ^{Footnote 20} it is much more about what states do for influence than about how influence is actually achieved.

Two assumptions are particularly integral to soft-power policies and scholarship. The first is that an individual is likely to look more favourably on a foreign country after getting to know its language and way of life: familiarity and personal contact are expected to engender affinity. This underlies the allocation of state funding to exchange programmes and language/cultural institutes, ^{Footnote 21} as well as expectations that an internationalised higher-education sector will produce soft-power dividends. ^{Footnote 22} The second assumption is that certain kinds of journalistic output are likely to produce more positive public evaluations of particular states. Opinions vary as to what kind of journalism will generate the desired effect: the UK government believes it benefits from the BBC’s reputation for impartiality, ^{Footnote 23} whereas other governments prefer their international broadcasters to take a more overtly patriotic line. Either way, confidence in the mass media’s ability to shape foreign public opinion is reflected in state financing for a multitude of internationally-oriented channels, news agencies, websites, and publications. ^{Footnote 24}

What evidence supports these two core soft-power assumptions? In the field of social psychology there is considerable, if qualified, support for intergroup contact theory—the idea that greater personal interaction with an outgroup reduces prejudice and hostility towards that outgroup, provided certain conditions are met. ^{Footnote 25} An inverse relationship between contact and prejudice at the individual level has been found across many contexts, so it seems reasonable to expect people to think more highly (or at least less negatively) about citizens from a particular foreign country, the more they interact with citizens of that country through travel, international exchanges, and so on. The extent to which affinity towards foreign individuals translates into to affinity towards the foreign state and its policies is less clear, however. ^{Footnote 26} It is worth noting the argument of Hugh Forbes, ^{Footnote 27} who suggests that contact at an individual level can ultimately exacerbate tensions at an aggregate level, because contact leads to assimilation, which may be resented by sections of the community being assimilated.

Evidence to support the assumption that the media shape public opinion about foreign states likewise exists without being clear cut. One recent study found a weak but positive relationship between China’s “media footprint” in African countries (i.e., Chinese involvement in local media environments) and the likelihood of those countries’ citizens looking favourably upon Chinese influence. ^{Footnote 28} However, the impact of media content was hard to assess, since the media footprint variable was based on Chinese telecom investment, as well as the presence or absence of several Chinese news outlets. In earlier research, William Gamson’s classic focus-group study of how individuals construct meaning from the news suggested that media discourse was more influential than “popular wisdom” and “experiential knowledge” where foreign affairs (specifically, the Arab-Israeli conflict) were concerned. ^{Footnote 29} Paul Brewer and colleagues found that students evaluated Mexico and Columbia more highly following exposure to stories that framed those countries as American allies in the war on drugs. The authors concluded that “a clear story—a frame that explains what issue is at stake and on which side of the issue the foreign nation stands” could shape how a mass audience judges another country. ^{Footnote 30} However, the connection between favourable/unfavourable media coverage and favourable/unfavourable public opinion about a country does not seem to hold in all circumstances. Wayne Wanta and colleagues observed a correlation between negative coverage of countries in American newscasts and low approval ratings of those countries among the American public—yet they observed no association between positive coverage and positive public perceptions. ^{Footnote 31} Another study found that measures of media attention explained less variation in opinions about West Germany than having German friends or relatives, German ancestry, and experience of travelling to Europe. For perceptions of East Germany, age was the strongest predictor of negative views. ^{Footnote 32} Some scholars have argued that images of enemies may be particularly stable and inertial elements in international belief systems and resistant to change in the face of dramatic turnarounds in events. ^{Footnote 33}

Thanks to the internet, individuals today can access a greater range of news sources than ever before. If tracing media influence on perceptions of other countries was challenging in previous decades, it is even harder now. It has been suggested that the media’s potential to direct public opinion may be diminishing as people acquire greater scope for selective exposure and general news avoidance. ^{Footnote 34} Yet the lack of firm evidence regarding the media’s ability to shape public opinion on foreign affairs has not halted the expansion of international state-funded broadcasters. Nor has it assuaged Western anxieties about the efforts made by authoritarian states to sway the thinking of international audiences.

Persuasion and Russian Foreign Policy

Russia (and, to a lesser degree, China) stands accused of “hijacking soft power” in an assault on democratic values. ^{Footnote 35} Rather than aiming for genuine persuasion, Russia’s leadership is said to be trying to “pollute the information space, increase polarization and undermine democratic debate,” ^{Footnote 36} creating so much confusion that people conclude “nothing is true” and “everything is possible.” ^{Footnote 37} The Russian government’s (mis)use of mass media in foreign policy became a matter of heightened Western concern in 2014, when Crimea was annexed and conflict broke out in Ukraine’s eastern Donbas region. During these events Russian state-controlled TV channels were repeatedly observed using “outright fakery” to support a narrative that absolved the Russian side of wrongdoing and painted Ukraine’s newly-installed pro-Western authorities as brutal and illegitimate. ^{Footnote 38} Notorious examples of faked Russian news include the alleged crucifixion of a toddler by Ukrainian troops, ^{Footnote 39} multiple far-fetched theories of how Ukrainian forces might have downed the airliner MH17, ^{Footnote 40} and claims that Ukrainian schoolchildren were being taught to kill bullfinches because the birds share the colours of the Russian flag. ^{Footnote 41} Several organisations now exist to identify, record, and “debunk” fabricated news stories in the Russian media. ^{Footnote 42}

Influencing public opinion abroad via the media is in fact a longstanding ambition of Russian policymakers that predates the crisis in Ukraine. It is mentioned as an explicit goal in all four of the Foreign Policy Concepts that Russia has issued since 2000 (these keynote documents summarise Russia’s international priorities). A significant change, however, is observable in the language and logic underlying this desire for media influence. Back in 2000—when Vladimir Putin had just assumed the Russian presidency for the first time—the primary stated aim was to “form a positive understanding of Russia abroad and a friendly attitude towards it.” ^{Footnote 43} By 2016 all references to friendly attitudes had disappeared; the emphasis now lies on “countering threats to information security.” ^{Footnote 44} The latest Russian Doctrine of Information Security complains about foreign states using information technologies to “undermine the sovereignty, the political and social stability and the territorial integrity of the Russian Federation and its allies.” ^{Footnote 45} A desire to counteract these alleged attempts to weaken Russia is at the heart of Russian communication policy. In Moscow, information is perceived as a weapon that the West has used effectively against Russia—to inspire the so-called coloured revolutions, for example. ^{Footnote 46} By the same logic, Russian state media are regarded as weapons for the defence of Russian interests both at home and abroad.

Disinformation and “whataboutism” undoubtedly feature strongly in Russian state-sponsored media content, but from this it would be wrong to conclude that the Russian leadership has no interest in persuading international audiences and seeks only to “dismiss, distort, distract and dismay.” ^{Footnote 47} For years, a highly consistent narrative has run through the content of Russian state-controlled media and official statements. The narrative problematizes American or Western “hypocrisy” and “interference”; blames these traits for global instability; and advocates a “multipolar” world as the optimal solution, in which non-Western states such as Russia would balance American power. ^{Footnote 48} The consistency of this narrative would be both unnecessary and unlikely if the Russian leadership had no desire for the narrative to be taken seriously. Falsifications are used to support the narrative—but they are not intended to turn everyone listening into nihilists for whom nothing is true.

Policy documents also indicate that Russia has had long-term goals vis-à-vis international audiences that extend beyond disruption, and this is particularly true in the case of Ukraine. The goal of helping Russian “diasporas” and “compatriots” abroad to preserve their “ethno-cultural identity” and “ties to the historic motherland” is explicit in the Foreign Policy Concepts of 2008, 2013, and 2016. Ukraine was highlighted as a priority partner, which the Russian leadership wanted to draw into “deeper processes of integration.” ^{Footnote 49} Essentially, the Russian leadership was hoping most Ukrainians would fall into line with its preferred narrative about their identity—as a population with historic, linguistic, and cultural ties to Russia that ought to be preserved in new structures of regional alliance. With this aim in mind, they made every effort to preserve channels of communication from Russia to the mass Ukrainian audience, but outcomes ultimately fell well short of their aspirations.

Ukrainian Attitudes towards Russia Before and After the “Revolution of Dignity”

At least until 2014, attitudes towards Russia were one of the main lines of cleavage in Ukrainian politics. ^{Footnote 50} An extensive area-studies literature addresses the fluctuating but fairly even divide in Ukrainian public opinion about relations with Russia that existed during the first 20 or so years of Ukraine’s independence. This literature, based on nationally representative surveys, is almost completely detached from research on soft power and persuasive cross-border communication. Table A1 lists almost a dozen relevant studies, along with the explanatory variables that were investigated and linked (or not) to variation in views (please note that all the tables mentioned in this article can be found in the Supplementary Materials). ^{Footnote 51} The general consensus was that “the combination of where one lives, what language(s) one speaks, what one’s ethnic identity is, and what religious group one is in tells us a great deal about mass attitudes in Ukraine” where Russia is concerned. ^{Footnote 52} Speaking Russian rather than Ukrainian, living in an Eastern region rather than a Western one, identifying as ethnic Russian and Orthodox (sometimes just religious), and being less well-off and older were the factors most frequently associated with pro-Russian (and correspondingly anti-Western) sentiments. However, there was some disagreement as to whether region of residence had an autonomous effect, or whether observed regional variations were due to the “compositional” influence of linguistic, ethnic, and religious differences. ^{Footnote 53} Another point of debate was how to treat variables that do not lend themselves to straightforward categorisation; in Ukraine this applies to region, language and ethnicity. It is striking that media consumption was not included as a predictor in any of the listed studies—despite use of Russia-based media having been widespread in Ukraine for a long time and television being considered a major influence on geopolitical imaginations. ^{Footnote 54}

The studies in table A1 are all based on fieldwork that predates “Euromaidan” and the “revolution of dignity” (revolyutsiya hidnosti) that ousted Viktor Yanukovych from the Ukrainian presidency in 2014. During and after the Euromaidan protests, Russian officials and state media harshly criticised the Ukrainian demonstrators and depicted them as hooligans working to a Western agenda. ^{Footnote 55} Some Ukrainians apparently sympathised with this point of view: in one nationwide poll from October 2014 (which excluded Crimea), around 30 percent of respondents said Euromaidan had been a “planned coup d’état,” against 38 percent who saw it as citizens uniting to defend their rights and 17 percent who said it was a spontaneous popular protest. ^{Footnote 56} Support for the coup d’état interpretation was predictably higher in the East and South than in the West and Centre. However, any putative “persuasive” effect of the Russian state’s message in Ukraine was offset by indignation and anger that the message and subsequent Russian actions also provoked. In a nationwide poll from June 2015 (again excluding Crimea), 60 percent of respondents said their opinion of the Russian media had deteriorated since the start of the year. ^{Footnote 57} Biased coverage of Ukraine in the Russian media became a prominent news story in its own right, reported and criticised by many popular Ukrainian channels and publications. A journalist from a leading Ukrainian TV channel even managed to hand an “Oscar” to a Russian correspondent live on air—for the “lies and nonsense” propagated by Russian state TV. ^{Footnote 58} In 2016 it was estimated that only around 5 percent of Ukrainians were continuing to watch Russian TV for news on a weekly basis; ^{Footnote 59} roughly 80 percent of Ukrainians said they had a negative or very negative view of the Russian leadership; ^{Footnote 60} and the long-held Russian ambition of drawing Ukraine into regional integration initiatives has entirely disappeared from the agenda.

From Soft Power to Linkage and Strategic Narrative Reception

If we return to thinking about how international persuasive communication is theorised, it should be clear why the soft-power framework is problematic. In Russia, Ukraine, the United States, and Europe, state-led international communication is being discussed in adversarial terms. This reflects the reality that government initiatives to persuade and attract were never quite as softly non-disruptive as their proponents claimed, regardless of the states involved. Any communication aimed at persuasion has a competitive dimension, because persuasion only has meaning in contexts where opposing views are vying for acceptance. Likewise, “attraction” only really matters when alternative centres of gravity are available. Nye underplays the competitive side of his soft-power concept, insisting that “it is a mistake to see public diplomacy simply in adversarial terms … often there can be gains for both sides.” ^{Footnote 61} But his much-cited assertion that foreign-policy success depends on “whose story wins” betrays a contradictory perspective that may be closer to reality: if one side wins, there surely has to be a losing side too. ^{Footnote 62} It is fear of being undermined by geopolitical rivals that has driven the Russian government’s intensification of propaganda, ^{Footnote 63} and now a feedback loop is emerging, with numerous Western politicians and commentators demanding countermeasures against the threat of defeat by Russian (dis)information. ^{Footnote 64}

The competitive dimension of persuasive communication is acknowledged in recent research that critically reworks the soft-power concept. Valentina Feklyunina, for example, proposes a ‘social constructivist take,” in which the soft power of State A vis-à-vis State B depends on (1) how widely narratives of collective identity projected by the former are accepted or resisted in the latter, and (2) how much influence the receptive audiences have over policymaking. ^{Footnote 65} This interpretation highlights the importance of the discursive context in which contestation between alternative narratives occurs. Alister Miskimmon, Ben O’Loughlin, and Laura Roselle go further in using the concept of narrative to explain how non-coercive influence works in international affairs. They build an innovative analytical framework around the concept of “strategic narrative”—an assemblage of messages through which a state tries “to construct a shared meaning of the past, present and future of international politics” and thereby shape others’ behaviour. ^{Footnote 66} A narrative can be understood as the accentuation and emplotment of particular problems or turning points in a way that indicates both causation and a normatively desirable resolution. Miskimmon, O’Loughlin, and Roselle contend that narratives projected strategically can exert a dual kind of power over behaviour: they may convince rational actors to act (consciously) in a particular way, while simultaneously constraining identities, understandings of the international system, and thus the longer-term (subconscious) formation of interests.

The strategic-narrative framework has the additional advantage of being relatively straightforward to operationalise for a study of reception—more so than the vague notion of attraction on which the soft-power framework is based. The key elements of a particular strategic narrative (issues problematized, claims of causality, and solutions advocated) can be identified from official statements and state-funded media content, then used as the basis for carefully designed questions, which are posed to different target audiences to assess whether they find the narrative convincing.

The strategic narrative framework is therefore adopted to structure the following empirical analysis of persuasive communication in Ukraine. Three expectations derived from the earlier discussion of soft power are tested; namely:

Three expectations derived from the area-studies literature on Ukrainian public opinion are also investigated:

Before proceeding to the research itself, a few important comments about these hypotheses are in order. First, the hypotheses should not be interpreted as claims of mono-directional causality. Variations in media use, personal connections, language preferences, and church attendance are expected to predict variations in support for the Russian and Ukrainian strategic narratives, but only in the statistical sense. Varying support for the narratives might equally predict varying patterns of media use, personal connections, and so on. The problem of assessing causal direction in the hypothesised relationships will be discussed in the concluding sections.

Second, Hypotheses 4, 5 and 6 have been formulated in a way that attempts to correct as much as possible for problems of endogeneity and categorisation that were not always recognised in previous research on Ukrainian public opinion. The response variables in the present investigation are measures of support for competing narratives. Attitudinal variables are not, therefore, used as predictors, because attitudinal variables would likely be endogenous to (i.e., alternative indicators of) support for the narratives of interest. Dislike of Ukrainian President Petro Poroshenko, for example, would probably correlate with any measure of support for Russia’s strategic narrative—but it would not meaningfully explain acceptance of the narrative since it constitutes acceptance of the narrative. A similar problem of endogeneity would arise with measures of identity that have been used in previous studies. Ukrainian survey respondents are often asked to place themselves into neat categories: “ethnic Russian” or “ethnic Ukrainian” (sometimes “mixed”); “Ukrainian-speaker” or “Russian-speaker” (sometimes “speaks both”); “Orthodox,” “Greek-Catholic,” or “no religion,” and so on. Yet ethno-national identities in much of Ukraine are characterised by “ambivalence and instability,” ^{Footnote 67} while many individuals “constantly switch from one language to another and often within the same sentence” without even registering whether their words are Russian or Ukrainian. ^{Footnote 68} Many of those who describe themselves as Orthodox or believers never attend church. The ethnic, linguistic, and religious self-categorisations achieved in surveys are not always, therefore, reliable indicators of variation in culture, heritage, or everyday practices. They may just as much reflect the individual’s acceptance of various ideas about national belonging that circulate in the public sphere. For instance, it is sometimes said that speaking or knowing Ukrainian is a vital part of being Ukrainian—and if a citizen believes this, it might increase their inclination to say she speaks both even if she predominantly communicates in Russian. Meanwhile, the Russian government has long projected the idea that any individual who speaks Russian or has some ancestral ties to Russia naturally belongs in a sense to the Russian homeland. ^{Footnote 69} By these criteria, a huge number of people living in Ukraine qualify to be Russian compatriots if they wish—so whether or not they describe their identity as Russian (fully or partially) reflects the appeal of Moscow’s strategic narrative relative to other, more local (and not necessarily strategic) narratives about identity. This is why the hypotheses in this study do not rely on ethnic, linguistic, or religious self-identification, but rather on the less-abstract measures of birthplace and habitual practices. The language hypothesis is formulated deliberately around passive language ability (TV viewing) instead of active language ability (speaking), because exposure to narratives depends more on the ability to understand than to speak fluently. Clear categorisation between Russian and Ukrainian also makes more sense in the context of television, where surzhyk (mixing of Russian and Ukrainian) is less common than in interpersonal communication. The hypothesis about religion focuses specifically on attendance of the Orthodox Church of the Moscow Patriarchate because this church has lately been described as one of the Kremlin’s soft-power “tools.” ^{Footnote 70} A regional effect is not hypothesised at all because the data used in this study come from only one Ukrainian region. The demographic variables found to be influential in some previous research (age, income level, education level, settlement type, gender) are included as controls.

A final comment, important for the subsequent discussion, is that most of the hypothesised predictor variables (media use, diverse personal connections, TV language preferences and religious attendance) can be thought of as types of linkage. In their book about transitions from authoritarianism, Levitsky and Way write that linkage is “rooted in concrete ties—networks; organizations; and flows of people, information, and resources—among states.” ^{Footnote 71} These ties, they argue, serve as transmitters of international influence. My study investigates variables related to flows of people and information, i.e., the social and communication aspects of linkage.

Levitsky and Way were not concerned with the reception of strategic narratives. In fact, they explicitly set aside the possibility of linkage having “ideational” mechanisms, focusing rather on how it affected the incentives of elite actors. Their work details how linkage to Western countries created incentives for democratisation in hybrid political systems. While acknowledging that “not all linkage is Western,” ^{Footnote 72} they operationalise the concept only in relation to Western states: they measure it by the extent of trade with the United States and EU states (economic ties); the proportion of citizens travelling to or living in the United States and EU states (social ties); international voice traffic and internet access per capita (communication ties); and membership in the Organization of American States or eligibility for EU membership (intergovernmental ties). These are all country-level measures, not individual-level ones.

My study therefore takes Levitsky and Way’s conceptualisation of linkage in a new direction, adapting it for the purpose of the problem in hand. Here, the focus is on individuals rather than states; economic and intergovernmental aspects of linkage are set aside; and linkage is operationalised in relation to Russia as well as democratic Western countries. Yet the basic idea of linkage is helpful because it provides a way of looking at persuasive influence that is not centred on particular resources (policies, values, media, etc.) generating advantageous attraction, but on individual patterns of behaviour that allow ideas to cross borders—with potentially diverse consequences. The implications of this more audience-focused perspective are elaborated in the concluding sections.

Reception of Russia’s Strategic Narrative in Ukraine: An Empirical Analysis

The data used here come from an original survey conducted between February 6–18, 2016 by the market research company TNS Ukraine. The sample of 1,000 respondents was representative of the adult population of Odesa Region. ^{Footnote 73} The survey was restricted to a single region (oblast) partly due to budget limitations, but also in order to set aside the regional-effect question while interrogating sub-regional, individual-level, explanatory variables more closely—particularly media use and personal ties to Russia. Odesa Region was selected because it belongs to the south-eastern half of Ukraine that voted for relatively pro-Russian candidates and parties in Ukrainian elections prior to 2014, yet it is also diverse in terms of its economy and linguistic composition. It has substantive interest in its own right, as it contains Ukraine’s third-biggest city and is part of the territory named Novorossiya (New Russia) by Putin and Russian nationalists. ^{Footnote 74} In 2014 Odesa was the scene of fatal clashes between supporters and opponents of the Euromaidan movement. Over 40 pro-Russian activists died in May that year when Odesa’s Trade Unions building was set alight during the unrest, an event described as a “massacre” by Russian state television. ^{Footnote 75} Ukrainian sociologists have since then identified Odesa Region as one where Russian propaganda is “most effective.” ^{Footnote 76}

The survey questionnaire had three sections. The first section included questions about respondents’ socio-demographic background, personal ties to Russia and other countries, and how often they discussed issues of international relations (for the frequency distribution of variables based on these survey items refer to table A2).

The second section of the survey pertained to news consumption. Respondents were asked to specify the language in which they preferred to watch TV programmes and to estimate the time they spent consuming news and current affairs via television and the internet (again, refer to table A2). They were also asked to name all the TV channels, radio stations, publications and websites they used to follow the news “during a normal week.” Responses were elicited via unaided recall rather than presenting respondents with a list, so in aggregate they can be understood as “mindful” news repertoires rather than “total” news repertoires. ^{Footnote 77} Assessing media use though self-reports is notoriously problematic, but a list-based approach is now widely used and the open-ended list approach has been found to perform better than other techniques as a measure of true exposure. ^{Footnote 78}

Each respondent’s reliance on Russian sources was operationalised by dividing the number of mentioned Russian sources by the total number of mentioned news sources. This gives a rough estimate of the importance of Russian sources in the respondent’s regular news media repertoire. ^{Footnote 79} However, the question of whether a news source is Russian is not always as straightforward as one might think due to partnerships between Russian and Ukrainian media companies. ^{Footnote 80} Moreover, Russian sources vary greatly in their degree of autonomy from the Russian authorities. Therefore, two variables are calculated, one based on a narrow definition of Russian media (the sources most clearly linked to the Kremlin), ^{Footnote 81} and one based on a broader definition (including more autonomous and commercial sources). ^{Footnote 82} In total, 14 percent of respondents reported using at least one of the “narrow definition” Russian sources, while roughly 34 percent of respondents reported using at least one of the “broad definition” Russian sources. Reliance on Russian sources was generally quite low (refer to table A3). Even among the minority of respondents who used one or more Russian news source (whether narrowly or broadly defined), these sources rarely constituted more than one-third of individual news media repertoires; respondents used more Ukrainian sources than Russian ones.

The third section of the survey questionnaire contained a series of questions intended to gauge support for the conflicting strategic narratives projected by the Russian and Ukrainian leaderships. Before formulating the questions, the strategic narratives themselves were traced by methodically analysing statements and speeches of the president and foreign minister of each country with the CAQDAS software Atlas.ti. All statements published on the Russian and Ukrainian presidential and Foreign Ministry websites between October 1 and mid-December 2015 were included in the analysis. Also studied were commentaries broadcast during the same period by Dmitriy Kiselev, presenter of the weekly news show Vesti Nedeli on Russian state TV, who is considered a major spokesperson for the Russian strategic narrative; there is no journalist who plays an equivalent role in Ukraine. Coding focused on three principal dimensions of the statements and commentaries which are derived from the definition of narrative given in the previous section: (1) the definition of problems in current affairs, (2) claims about the causes of those problems, and (3) solutions advocated as normatively desirable.

During the studied period certain topics were covered by both narratives from contradictory perspectives, and these became the focus of the survey questions. One set of questions asked respondents to state their level of agreement with alternative problem definitions from the two narratives, for example, “The USA violates the sovereignty of other countries,” or “Russia is trying to destabilize Ukraine.” A second set of questions asked respondents to choose between contradictory causal claims taken from the two narratives, for example, “Russian support for separatists” versus “Ukrainian unwillingness to give Donbas special status” as explanations for why the conflict in eastern Ukraine had not been resolved. A third set of questions asked respondents to choose between contradictory solutions advocated by the rival narratives which might “improve the situation in the world,” for example, the United States and Europe continuing sanctions against Russia, or the United States and Europe cooperating more closely with Russia (for a full list of these questions and the frequency distribution of responses, refer to table A4 and table A5 in the Supplementary Materials).

An exploratory factor analysis was run on the problem definition set of survey items. ^{Footnote 83} A two-factor solution was found to account for 74 percent of total variance (refer to table A6). All the Russian problem definitions loaded onto Factor 1 and all the Ukrainian problem definitions loaded onto Factor 2, although one Russian problem definition (“Countries of the West and Europe are losing interest in solving Ukraine’s problems”) cross-loaded and was therefore discarded from the analysis. The two factors were negatively correlated with a coefficient of -0.25. ^{Footnote 84}

A second exploratory factor analysis was run on the causality and solution sets of survey items. ^{Footnote 85} A single-factor solution was found to account for 48 percent of total variance (refer to table A7 in the Supplementary Materials). ^{Footnote 86}

The results of these factor analyses are consistent with the existence of two negatively-correlated latent variables—support for the Ukrainian strategic narrative and support for the Russian strategic narrative—underlying responses to the narrative-based survey questions. For subsequent analysis, these latent variables are operationalised by standardizing the respondents’ scores on each narrative-based survey item, then calculating the unweighted mean of the standardized scores across the items contributing to each factor. ^{Footnote 87} This produces three measures of narrative support: (RV1) agreement with the Ukrainian problem definitions (based on Factor 1 from table A6); (RV2) agreement with the Russian problem definitions (based on Factor 2 from table A6); and (RV3) agreement with the Ukrainian causal attributions and solutions rather than the Russian causal attributions and solutions (based on Factor 1 from table A7). Summary statistics for these three measures are presented in table A8.

Table A9 presents models that regress the three measures of narrative support (RV1, RV2, and RV3) on the hypothesized explanatory variables. The results support Hypothesis 1 that individuals will support the Russian narrative more (and the Ukrainian narrative less) the more they rely on Russian news sources. Reliance on Russian media—whether narrowly or broadly defined—is among the strongest predictors of disagreement with Ukraine’s anti-Russian problem definitions (Models 1 and 2), causal claims, and advocated solutions (Models 5 and 6). The association between reliance on Russian media and supporting Russia’s anti-Western problem definitions is somewhat weaker (in Model 3 it is not significant; in Model 4 it is significant only at the 0.1 level) although the coefficients point in the expected direction. It is worth noting that the association between disagreement with the Ukrainian narrative and using Russian news media is not driven solely by the narrowly-defined sources (i.e., Russia’s main federal channels and state news agency RIA Novosti). The more broadly-defined Russian sources (which include the popular social networks Vkontakte and Odnoklassniki) seem to have an effect of their own that can be observed in “time spent following news online” having larger coefficients in Models 1 and 5 (where the narrow definition of Russian media was used) than in Models 2 and 6 (where the broad definition was used).

Hypotheses 2 and 3 postulated that personal ties to Russia and Western countries (communication with friends and relatives there, regular travel and experience of life there) would predict reception of the rival narratives. Regular travel to Russia certainly seems to be significant: it is a strong predictor of disagreement with the Ukrainian narrative and agreement with the Russian narrative across all six models. However, experience of having lived in Russia was found to have no significant effect. Regular communication with friends or relatives in Russia had a small but significant positive effect on the respondents’ support for Russia’s anti-Western problem definitions, as well as the Russian causal claims and advocated solutions, although it did not significantly affect support for Ukraine’s anti-Russian problem definitions. Regular travel to Western countries was associated with stronger support for Ukraine’s anti-Russian problem definitions, yet had no significant effect on support for Russia’s anti-Western problem definitions nor on respondents’ views about causality and solutions. Experience of having lived in the West was associated with stronger support for the pro-Western Ukrainian narrative (problem definitions, causality and solutions) but no effect was observed on acceptance of Russia’s anti-Western problem definitions. Communication with friends and relatives in the West, on the other hand, reduced support for Russia’s anti-Western problem definitions while having no observable effect on support for Ukraine’s anti-Russian problem definitions nor on views about causality and solutions. Overall, therefore, personal cross-border connections do seem to have influence in the expected direction—but different kinds of connection have different levels of ‘effect’ depending on the narrative elements under study.

Hypothesis 4 predicted that individuals born in Russia would be more inclined to support Russia’s strategic narrative over Ukraine’s strategic narrative. This hypothesis was not supported: country of birth had no effect on support for the different narratives in any of the models.

Hypothesis 5 was strongly supported by all the models. Respondents who preferred watching TV programmes in the Russian language demonstrated substantially lower support for Ukraine’s anti-Russian problem definitions, causal claims and advocated solutions, as well as greater support for Russia’s anti-Western problem definitions, compared to respondents who preferred Ukrainian-language programmes or had no preference between Russian and Ukrainian. Interestingly, no differences were observed between the respondents who preferred Ukrainian only and those who had no preference between Ukrainian and Russian. The Russian-only TV viewers differed markedly from both.

Hypothesis 6 predicted that regular attenders of the Russian Orthodox Church of the Moscow Patriarchate would support the Russian strategic narrative more than non-attenders, and this was borne out by Models 1–4. ROCMP attendance was a significant predictor of disagreement with the anti-Russian problem definitions from Ukraine’s strategic narrative and of agreement with the anti-Western problem definitions from Russia’s strategic narrative. However, church attendance had no observable effect on support for the causal claims and solutions from the two narratives.

Among the socio-demographic and other variables investigated, age, income, and the frequency of discussing international issues were the only ones to have significant effects on support for the different narratives. Older age corresponded with lower support for the Ukrainian strategic narrative and higher support for the Russian strategic narrative in all six models. The more often respondents discussed foreign affairs, the more they tended to support the anti-Russian problem definitions, causal claims, and advocated solutions from Ukraine’s strategic narrative (although no effect was observed on their support for Russia’s anti-Western problem definitions). The reason for this association is not entirely clear. It may suggest that political engagement (reflected in frequency of political talk) is higher among more patriotic respondents who have negative views of Russia. Alternatively, the current social climate in Odesa Region may be more conducive to expressing anti-Russian opinions in conversation. The findings for income are contradictory. Models 5 and 6 suggest that greater wealth is associated with more support for Ukraine’s strategic narrative over Russia’s, yet Models 3 and 4 found the low- and middle-income groups to be more likely than the lowest- and highest-income groups to support Russia’s anti-Western problem definitions, while income had no significant effect in Models 1 and 2 (respondents are not necessarily open about their incomes; this might have affected results). The variables gender, education level, and settlement type, as well as several plausible interaction effects, were tested but not found to be significant.

Linkage, Strategic Narrative Reception and Foreign Policy “Success”

By sticking to Russian-language TV, travelling regularly to Russia, communicating with acquaintances in Russia, using Russia-based news sources, and attending a church headquartered in Moscow, some Ukrainian citizens maintain a connection to Russia, and these citizens express stronger than average support for the Russian state’s narrative about international affairs. On one level, this finding is unsurprising: it is consistent with the basic arguments of prior area-studies work that religion and language matter for Ukrainian attitudes towards Russia, as well as widely held (but previously untested) assumptions about Russian media influence.

The new step proposed here is to interpret these expected empirical relationships through the lens of linkage and ongoing processes of cross-border communication. Previous scholarship frequently pointed to identity-related factors—“native” language, “ethnicity,” and declared religion—as explanations for foreign-policy preferences and attitudes towards Russia among Ukrainians. But it tended to treat these factors as if they were static and exogenous; it said little about the individual communication-related activities that are integral both to the maintenance of identities and to the reception of competing ideas about international affairs. A relevant finding to emerge from the present analysis is that birthplace—a possible precursor to national identity—was insignificant for narrative reception among the studied population. Yet variables based on regular behaviours—church attendance, media consumption and interpersonal contact—did matter.

What can be said about causality in the relationship between linkage and strategic narrative reception? As already stated, my intention is not to suggest that the predictor variables of linkage cause variation in attitudes in a one-directional, linear fashion. Linkage should not be understood purely as a proxy for exposure, whereby the “stimulus” of greater exposure to a given narrative generates an automatic “response” of greater agreement. An obvious objection to the stimulus/response view is that the relationship over time between exposure to and support for a narrative seems likely to be mutually reinforcing. If an individual sympathises with one narrative more than another, she may be less likely to exclude the favoured narrative from her communicative environment, while giving less attention to contradictory perspectives. Over time, accustomisation to the favoured narrative might thereby have a structuring, limiting effect on how the individual understands the international system, ^{Footnote 88} rendering certain ideas or values as taken for granted, but this should be regarded as an “ongoing, dialectical process” rather than an “aggregation of isolated causal collisions.” ^{Footnote 89}

One can also think of ways besides long-term narrative exposure in which linkage is implicated in the process of narrative reception. Forms of linkage such as religious participation, cross-border travel, and close personal relationships generate practical and emotional reasons to value friendly ties between one’s own state and another. Imagine someone living in Ukraine who has many friends and relatives in Russia and likes to visit them; or someone who happens to enjoy Russian-made, Russian-language entertainment shows much more than Ukrainian-made, Ukrainian-language ones; or someone who prays for the fortunes of the Moscow Patriarchate in Ukraine because they consider it the “true” authority of faith. All such people might fear seeing something subjectively valuable to them disrupted by antagonistic Russia-Ukraine relations, which make cross-border flows of people and information more awkward. Correspondingly, they have more reason to support narratives that prioritise the restoration of bilateral ties, and to object against narratives that present the restoration of ties as a remote or undesirable prospect. In a sense, the behaviours that constitute linkage are not just conduits for ideas about values, they have value of their own which is liable to inform the reasoning of the people involved. Similarly, linkage is not just a conduit for narratives that establish a discursive divide between “us” (the collective Self) and “them” (the Other). Linkage consists of social behaviours through which individuals live out and reify their belonging to collectives that transcend state borders—be it the Orthodox Church, the “Soviet people,” or the “Russian world.” Given these complexities, linkage cannot be understood as a set of independent variables that generate an effect on attitudes. In the words of one non-positivist scholar, “the conceptualization of factors as independent forces only impedes understanding of both their dynamic interactions and their cumulative significance over time for the subjects we are trying to understand.” ^{Footnote 90}

Some might regard the proposed idea of linkage as quite compatible with Nye’s concept of soft power, given that the latter does stress the importance of interpersonal contact and channels of communication. To the extent that soft power has become a “nebulous” catchphrase, ^{Footnote 91} applied to all kinds of non-military forms of influence, there is some overlap. Yet the book that proclaimed soft power to be “the means to success in world politics” began by stating that “soft-power resources are assets that produce attraction,” and then identified U.S. international broadcasting to be one such resource that ought to receive greater investment. ^{Footnote 92} The implication is that a reasonably straightforward linear relationship exists between the intensity of a state’s efforts to project a particular message and its prospect of achieving foreign policy goals. The crucial fact that a projected message is likely to elicit diverse responses in the context of competition from other messengers was therefore obscured, and the challenge of explaining diverse responses—as tackled in this article—was not prioritised.

It is not logical to assert that persuasive efforts will bring foreign policy success without first developing a fuller understanding of reception that encompasses all potential responses—not only the responses of targeted groups, but also those of competitors in the “market for loyalties.” ^{Footnote 93} State-led persuasive efforts are clearly capable of generating backlash and divisions in addition to support. State A might, for example, see its strategic narrative becoming more widely accepted among the elite or general population of State B. For some scholars, these “minds changed,” ^{Footnote 94} or “views held,” ^{Footnote 95} would already be evidence of soft power (i.e., success) in action. But at the same time, certain sections of the elite or population in State B (or C, or D) might reject State A’s narrative, increasingly resent its growing influence and seek to counter it. The outcome would thus be rising tension with potential for instability, which might be far from the goal State A had hoped to achieve. The problem mirrors that described by Forbes in relation to the contact hypothesis: an increase in inter-group affinity at the individual level does not necessarily mean conflict will decrease at the aggregate level. ^{Footnote 96} Success in projecting a narrative that resonates is one thing; success in achieving foreign policy goals may be quite another—so the fact that these issues are conflated so regularly is problematic.

In recent years both Western and Russian efforts to shape public opinion in foreign countries have drawn a strong backlash. The repressive measures Russia introduced domestically to counter the “threat” of Western democracy promotion are already well documented; ^{Footnote 97} its more active deployment of international broadcasters and pseudo-nongovernmental organisations in foreign policy is another aspect of the same response. Yet the Russian government’s own drive to promote narratives abroad has generated undesired outcomes too. Even before Euromaidan, Crimea, and the Donbas conflict, influential sections of the Ukrainian elite were reacting with vocal criticism to Moscow’s promotion of a “Russian world” encompassing Ukraine. Since 2014, the Ukrainian authorities have worked hard to curtail flows of information and people from Russia (direct flights have been banned, as have many Russian entertainment programmes and the cable transmission of Russian TV channels). Moreover, anti-Russian narratives are now the norm in mainstream Ukrainian journalism. Thus, despite extensive social and communicative linkage allowing wide dissemination of Russia’s narrative to the Ukrainian population, Russian strategic communication vis-à-vis Ukraine cannot be judged a success. Messages projected by the Russian government may have helped to polarise opinion in Ukraine, but polarisation was not, in fact, the original objective. Russia’s original strategic objective was to draw Ukraine consensually into structures of regional alliance, and this now seems like a very distant prospect indeed.

Reframing Debates about State-Led Persuasion in International Relations

I have identified a set of individual behaviours that predict variation in the reception of strategic narratives in Ukraine. I have adopted and adapted the concept of linkage to explain how these behaviours might play a role in narrative acceptance. Breaking from previous research on soft power, I have also pointed out that the acceptance of a strategic narrative by a target audience should not automatically be equated with foreign policy success if negative responses from other audiences, including rival states, are elicited as well.

The arguments developed here about the role of linkage are based on data from one Ukrainian region, and this raises questions about their broader relevance. It would be worth conducting similar studies elsewhere in Ukraine, for a start, to ascertain whether the observed relationships hold in other regions. The prospect of replicating the findings more widely within Ukraine would appear to be strong, given that language- and religion-related variables are already established as predictors of Ukrainian attitudes about Russia. In the present study, language use and religion were measured and interpreted in a slightly different way (emphasising concrete social or communicative behaviours rather than abstract identities) and media use and interpersonal ties were introduced as important aspects of linkage that had previously been neglected. Future studies could perhaps compare the behaviour-based variables used here with more traditional measures of religious, linguistic, and ethno-national identity, to see how they interact and which approach has the greater explanatory power.

Looking beyond Ukraine, individual-level linkage seems most likely to affect strategic narrative reception in contexts where history or geography have allowed it to develop over time. Strategic narrative reception would need to be explained by other factors in cases where the extent of individual social and communicative linkage is too limited for its role to be significant. Linkage between Russia and Ukraine is unusually extensive—there are not many states, even neighbouring ones, whose populations are so linguistically, culturally, and historically intertwined. It should be noted that the Russian news sources used by a substantial minority of Odesa Region residents are Russian domestic news sources; they were not designed to be international broadcasters for foreign citizens. In most parts of the world, consumption of news from foreign sources will probably not be as high as it is in Odesa Region. Dedicated international broadcasters that try to reach geographically and culturally distant populations often have relatively low audience shares (this applies to the Russian state channel RT in the United Kingdom, for example).

However, linkage as a factor in the persuasive process would be worth investigating in other parts of Eastern Europe and the former Soviet Union. For example, citizens in Moldova, Belarus, and Armenia face a similar choice of narratives that lead in different foreign policy directions—towards Russia and the Eurasian Union, or towards the European Union and the West. These citizens also maintain varying levels of personal, linguistic and cultural ties to different nearby states, just like Ukrainian citizens do. There is currently concern about the “influence of Russian propaganda” in the countries of Eastern Europe. Reframing this issue in terms of linkage and strategic narrative reception could help shift analysis away from a simplistic stimulus/response model of media effects, towards a more holistic appreciation of how views of international politics shape and are shaped by the social and communicative practices of individuals.

I set out to explain the reception of two strategic (long-term) narratives that are clearly geopolitical in nature—they describe heroes and antagonists on the international stage. The reception of messages that are less strategic and less geopolitical would probably be less affected by linkage. For example, the Russian government is accused of financing all kinds of media content that worked to Donald Trump’s advantage in the 2016 American presidential election (particularly conspiracy theories and leaks damaging to Hillary Clinton). Reception of this more tactical kind of messaging about specific domestic political issues lies beyond the scope of this study.

A general conclusion of the present article is that soft power should cease to be the default label for the study and practice of mass persuasion in international politics. Soft power implies that efforts to change political preferences among foreign populations are benign and unobtrusive; in reality they may provoke resentment and retaliation. Soft power implies that states get what they want by persuading foreign audiences, but this is questionable when persuasion or attempts to persuade are followed by backlash and division. The same resources that help attract some Ukrainians to Russia (i.e., church and media) repel other Ukrainians, because different audiences receive and respond to narratives in different ways. Any resource-based account of how persuasion works is therefore inherently problematic.

I have suggested that linkage is more than just a conduit for discourse. Cross-border travel and talk, media consumption, and religious participation are likely to affect an individual’s exposure to different narratives over time. However, their significance for narrative reception may also lie in the fact that an individual will resist narratives that threaten the cross-border activities and contacts which she values. This point is pertinent to the promotion of liberal democracy in culturally Russified areas via the media and information campaigns. Unfortunately, the narratives that promote liberal democracy are very often narratives that also depict Russia as a threat and interaction with Russia (Russian influence) as generally undesirable. If pro-democracy narratives in the media were less thoroughly geopoliticised, they might go down better among individuals who have strong ties to Russia which they value and wish to maintain.

Although I have emphasised the risk of backlash associated with state-led efforts to persuade foreign audiences in a competitive discursive environment, I do not mean to imply that such efforts should not take place. It is inevitable that conflicting strategic narratives will be projected in the course of international politics. From a normative viewpoint, one may consider it right to project narratives promoting democracy irrespective of how non-democratic governments react. However, there should be greater recognition of the fact that the reception of a strategic narrative will sometimes have little to do with how skilfully it is deployed, nor with how widely and actively it is disseminated, nor even with how faithful it is to the “truth.” Credibility is not an objective property of a source or message, but a receiver perception. ^{Footnote 98} Individuals will assess the credibility of strategic narratives against the yardstick of their existing views of the international system—views that are likely to have been shaped over time by the behaviours described here as linkage.