Monitoring of hate speech in Belarusian mass media: aims and methods

время чтения: 12 мин

Monitoring team

Monitoring is carried out by volunteers of «Journalists for Tolerance» initiative group.

The monitoring idea and concept: Aleh Razhkou and Andrus Klikunou.

The monitoring methods / expert analysis: Violetta Yermakova.


Violetta Yermakova

  • Master of Political Science (EHU, 2009), thesis on “Tolerance as a characteristic of the political culture of Belarusians”;
  • Coordinator on the issues of gender at East European  Network  of the Peace Organizations (EENPO) of the Global Partnership for the Prevention of Armed Conflict (GPPAC);
  • Jury member of the International Competition for the best publication on tolerance, integration, mutual respect and understanding in society, carried out by EENPO since 2010;
  • Research and publications:
    • “The Map of Tolerance”, 2011
    • “Gender Educational Program” (parts 8 and 10), 2013
    • Tolerant Journalism”, 2013
    • “The issue of the disabled people’s employment in the mirror of mass media”, 2014;
    • Member of “Journalists for Tolerance” initiative group;
    • ‘Radio Racyja’ correspondent.

Monitoring aim and objectives

The monitoring is aimed at assessing:

  • the scope of widespread manifestations of hate speech in Belarusian mass media,
  • the social groups it affects most of all;
  • the categories of mass media with frequent manifestations of hate speech in Belarus (national /regional, state/ independent);
  • the producers of hate speech in the media (the characters / the authors).

The monitoring data has been collected both for scientific and pragmatic reasons. The “Journalists for Tolerance” initiative members believe that the culture of creating ethic journalistic materials and the journalists’ feeling of social responsibility have significant development potential. And it is important to realize how this potential can be applied.

Various educational seminars and workshops on Ethic Journalism are regularly arranged in Belarus. The corresponding trainers require factual information on the most wide-spread manifestations of hate speech in Belarusian mass media to include them as illustrative cases in their presentations. Also, the trainers need to know for whom they should conduct such workshops, in order to have a noticeable effect in the national media field. Should the target group include journalists from the regions or the capital city, the employees of the state or independent mass media? Is it really important to work with journalists or do journalists only quote incorrect statements of their characters, being independent and unbiased themselves? Thus, the monitoring provides Ethical Journalism trainers and scholars with valuable information for their work.

On the other hand, the public importance of monitoring results is hard to overestimate.

Their publication helps to draw journalists’ attention to the topic of hate speech. The topic used to be marginal till the recent past. And it used to be discussed from the legal perspective, i.e. what publications can lead to suits against authors of arguable journalistic materials. The monitoring team made an attempt to shift the focus from the legal aspects to the ethic ones.

Any professional journalist possesses a bunch of refined and incredibly powerful communication tools. On the one hand, they can foster the increase of social tensions. On the other hand, they can contribute to resolving conflict situations.

The monitoring team decided to draw the media workers’ attention to this power and help them use it in smart and responsible ways.

Finally, the monitoring and its presentation are aimed at making the research visible for prospective monitors, who would like to join the team.

What is hate speech?

 It is not the case, when somebody speaks badly about another person. If a person speaks up negatively about someone who has stolen a wallet or about an official taking a bribe, it shouldn’t be considered as hate speech. However, derogatory statements in relation to someone, caused by the ethnic origin, gender, age, or sexuality make distinct manifestations of hate speech.

In other words, hate speech is the communication practice that generates and perpetuates discrimination. And the term ‘discrimination’ is not taken in its legal meaning here, but in the meaning that has been introduced in social studies and psychology. The significant difference is that lawyers, when talking about discrimination, view its outer side, the committed acts. Sociologists and psychologists pay attention to the internal conditions of behavior (attitudes, beliefs, stereotypes). Thus, in our understanding not only calls for actions (for example, to deprive someone of the right to education) are regarded as manifestations of hate speech, but also the spreading of negative stereotypes, the formation of negative images of social groups (e.g., describing them as unable to learn etc.) are considered this way.

Why is it important to investigate hate speech? The monitoring team believes in Sapir-Whorf hypothesis, according to which language determines our thinking. And then thinking determines behavior. Here is an example: an employer does not employ a Roma (discrimination, behavior), because he thinks «Gypsies are prone to theft» (thinking), and he thinks so because he gets the idea of a Roma from the media, where they are consistently mentioned in criminal chronicle (communication practice). Change of communication practice leads to change of attitudes (it’s a slow process, we should admit), and as a result of attitude change the probability of discrimination decreases.

So, at their discretion and according to their awareness, journalists can create a picture of the world in which the probability of discrimination in relation to a particular social group either increases or decreases. In fact, mass media have power to program the world of tomorrow to a certain extent. And the objective of our initiative group is to draw the journalists’ attention to the ‘magic wand’ in their hands.

Selection of monitoring data

It looks like this: six volunteers (not experts in hate speech, just journalists who have completed a special training course) read mass media Web-sites, select the journalistic materials that have manifestations of hate speech (the monitors have a list of such manifestations), and fill out the form that can be found at the end of this document. Then the collected materials are gathered together and checked by the person responsible for the monitoring methods and expert analysis.

There were analyzed 36 Belarusian media within the reporting period (see. Appendix No.1. List of mass media). The media had been chosen randomly.

Selection criteria: the monitored media outlets mainly deal with social and political problematic and have regularly updated Web-sites. The latter is a purely technical limitation, simplifying information processing. 2/3 of monitored media outlets are nationwide with broad audience and 1/3 of monitored media outlets are regional. The latter have been taken to see if there might be any difference in their performance in comparison with nationwide periodicals. A half of the monitored media are owned by the state. The other half of monitored media are independent.

The monitoring team agreed not to limit themselves to newspapers. Therefore, the Web-sites of some prominent TV and radio broadcasters were included into the list of monitored media, too.

The monitoring team was criticized by some fellow journalists for the ‘unrepresentative selection’ of materials. The critics claimed that the monitoring subjects did not make a miniature model of Belarusian media field. The monitors would be happy to learn the existing models of Belarusian media field, and it is likely that the team will be able to use them to improve the conducted research. Presently, the selection is done in an extremely simplified manner. However, it is quite sufficient for meeting the initially set objectives. (see above — «Monitoring aim and objectives» section).

The problem with the selection criticizing, in the monitors’ opinion, is not connected with the degree of representativeness, but in the fact that ‘the anti-rating list’ (i.e. a list of publications with hate speech) is published in the monitoring report. The criticizing journalists claimed that even harsher publications could be found in the media that didn’t enter the monitoring list.

The monitoring team admitted fair criticism and discussed ways to avoid it in the future (e.g., through broadening the monitoring list and the monitoring team at the general stage of the hate speech monitoring in the future).

Monitoring data analysis

The collected data were gathered in one table and calculated (see Appendix No.3. The Monitoring Table.)

Where did the numbers come from? The monitors’ challenge was not just to collect a database of materials containing hate speech in 36 media under review. They also needed to compare materials to determine which of them was more or less correct. And they needed to bear in mind that they were written on different topics, in different genres, and finally, that they differed in size.

For the comparison to be made possible, the monitoring expert analyzed the manifestations of hate speech in each article in question. At this point the numbers appeared, as the expert assigned numerical values to them. The thing is that hate speech can be of a different degree of harshness. 3 points were assigned for the most severe manifestations (marked in bold font), 2 points – for the moderate (marked in italics), and 1 point (unmarked) — for minor manifestations of hate speech (see the monitoring form below).

There were taken three scales of ‘Hate Speech’ harshness in the monitoring research – ‘Sova’ Center’s scale, “Xenomonitor” initiative’s scale, and the monitoring expert’s original ‘Tolerance of article’ scale. The marks from all three scales were summed for each article in question to make a rating list of ‘hate speech’ harshness manifestation.

However, it was not enough to have an objective and clear picture, since hate speech could set the message of the whole material, and could appear once only by chance. Journalists could give incorrect citations or they could produce hate speech themselves. These factors had to be considered, too. Therefore, it was decided to apply the increasing and decreasing amendments to the numbers received at the first stage. They were reflected in the table below:

Amendments Factors that influenced the perception of hate speech by the audience
Increase by 1.3 times Hate speech is typical of the whole material
Increase by 1.3 times The journalist produces hate speech or shows solidarity with those who produce it
Increase by 1.2 times Hate speech is common, but it does not capture all of the material
Increase by 1.2 times The journalist doesn’t express his attitude to hate speech in the words of his characters or his attitude is contradictory
Increase by 1.1 times Hate speech occurs no more than 3 times in the material
Decrease by 2 times The journalist expresses a negative attitude towards hate speech

Following the adjustment, the results for each monitored material under review could be compared to each other. (The monitoring team agreed to regard the figure as ‘the incorrectness indicator’.)

The results of these comparisons were published in the monitoring reports.

The monitoring doesn’t answer the following questions:

(1)     We can’t say, if there’s «a lot» or «little» hate speech in the monitored Belarusian media.

To do this we should compare our results with the results of, for instance, neighboring countries. And for such a comparison to be valid, it is necessary that the monitoring is conducted in several countries according to completely identical procedures. It is beyond the monitoring team’s capabilities nowadays.

We can only say what sort of hate speech is found in the monitored Belarusian media and give its qualitative description in the Final Evaluation Report (see below).

(2)     We can’t say if the use of hate speech has increased or decreased in the Belarusian mass media during the monitoring period. We can answer this question about those 36 media we’ve been monitoring, but definitely not about the whole country. The thing is that some journalists know that they are being monitored, and others know that they are not. And theoretically the observer’s effect could fire, since people tend to change their behavior to some extent, when observed.

(3) We can’t report on the percentage of hate speech containing publications. Technically, it was hard to do at the pilot stage. An interesting indicator could have appeared from this figure: the percentage of hate speech containing publications for different media. And mass media could be compared according to this indicator. We’re planning to include the marker at the general stage of ‘hate speech’ monitoring research in the future.

The questions our monitoring research gives answers to…

…are reflected in our reports (see Final Evaluation Report below).

The monitoring form

Name of the article
Date of publication
Genre (one answer only)
  • · a piece of news
  • · analytical article
  • · author’s column, blog
  • text of an organization’s declaration
  • ·interview with a public person
  • ·interview with an expert, expert commentary
  • ·press release or information message from an organization’s website
  • ·Separate phrase or quotation (not in the material but as a genre: aphorism, the phrase of the day)
  • ·readers’ questions and answers (consultations in the media, hotlines)
  • ·other


Hate speech location within the material (any answer option)
  • ·in the text of the article
  • ·in the headline
  • ·in the subheading
  • ·in the statements of the article’s characters (non-public figures) in the text of the article.
  • ·in the statements of the characters – public figures
  • ·in the statements of experts
  • ·in journalist’s questions and comments
  • ·in a citation from a document or press-release etc.
  • ·the material is in the featuring article or announced in the featuring article
  • ·hate speech is a highlighted (font, color, offset) fragment of the article
  • ·the article is incorrectly illustrated or has an incorrect caption
  • ·the article is accompanied by an incorrect joke, poem, proverb, or other text-illustration
  • ·hate speech in editor’s section
  • · other variants
Intensity of hate speech in the material (one answer)
  • · Hate speech is typical of the whole material (the number of manifestations of HS is greater than or equal to 1/2 of the text)
  • · HS is common, but does not capture all of the material (the number of manifestations is less than 1/2 the number of sentences)
  • ·HS is found in the material (in some words and phrases, no more than 3 times)
Subject of HS (one answer)
  • · article character(s)
  • · journalist
  • · both (including the indirect quotation and paraphrase)
Subject of HS: characters (if points 1 and 3 marked in the previous question)
  • · a politician
  • ·an MP
  • ·a representative of the NGO / initiative
  • · a representative of the church
  • ·an expert
  • ·law enforcement officers and employees of the judicial system
  • ·culture and art workers, «stars»
  • · ‘ordinary people’
  • ·classic culture representatives and historical figures (often in the form of quotations)
  • ·scientists and educators
  • ·other variant


The names and positions of the characters (if in the question about the subject first and third options were selected)
The author’s attitude to HS (one answer)
  • ·journalist doesn’t  express any attitude to hate speech manifestations (either no comments or neutral ones)
  • ·journalist expresses a negative attitude
  • · the journalist supports hate speech manifestations, expresses solidarity
  • · The journalist’s attitude to HT manifestations is contradictory (for example, partly supportive and partly disapproving)
Signs of selecting hate speech objects (all possible answers)
  • · civilizations, cultures
  • · race
  • ·supra-ethnic and supranational entities (the Slavs, the EU, etc.)
  • · understate entities (Caucasus and Caucasians, Western Belarus)
  • ·state as a substitute for the concept of «the people of the state»
  • ·ethnicity
  • ·citizenship
  • ·religion
  • ·sex
  • ·gender identity
  • ·sexual orientation
  • ·disabilities
  • ·other variants


The image formed by HS in the reader’s mind (all possible answers)
  • · negative “them”
  • · (hyper) positive “them”
  • ·negative “us”
  • ·(hyper) positive “us”
  • ·the image of differences (insurmountable, inborn, socially significant) between “us” and “them”
  • ·it’s difficult to determine which image is created in the material
Type of HS (the scale by “Sova” centre)
  • A. incitement to violence  in relation to a particular situation, indicating the object of violence; proclaiming violence an acceptable means in their articles, documents, etc .; including such abstract appeals as «Kill the Jews!»);
  • · B. incitement to discrimination, including in the form of generalized slogans;


  • · C. veiled incitement to violence and discrimination (promotion of «positive», historical or contemporary, examples of violence or discrimination, expressions like «it would do them good to…», «it’s high time …» and so on);
  • ·D. creating a negative image of a social group, not associated with specific charges but rather transmitted in the tone of the text or its parts;
  • ·E. Justifying historic cases of violence and discrimination (like «The Turks slaughtered Armenians in 1915 in self-defense»);
  • F. publications and statements that cast doubt on universally recognized historical facts of violence and discrimination (for example, the extent of the Holocaust, or the statement that «the Chechens were deported because they had sided with Hitler»);
  • G. Statement of deficiency (lack of culture, intellectual ability, inability to creative work) of a particular social group (such ideas as «disabled people do not want to work»);
  • H. Statements of historical crimes of a particular social group (such as «Muslims have always spread their faith by force», «Poles have always made plots against Russians»);
  • I. allegations of criminality in a particular social group (e.g., «Gypsies are thieves»);
  • · J. statements of moral shortcomings of a particular social group («Jews are hucksters», «Gypsies are liars» – to be distinguished from statements of cultural or intellectual inferiority);
  • K. discussions on the disproportionate superiority of a particular ethnic or religious group in material prosperity, representation in the government, the press, etc.;
  • L. charges of the negative impact of a particular social group on the society, the state («the erosion of national identity», «foreigners transforming Moscow into a non-Russian city», «Mormons undermining our Orthodox identity”);
    • M. naming social groups or their representatives in a derogatory or insulting context (including the criminal chronicle, or just mentioning an ethnonym);
    • ·N. appeals to prevent consolidation in the region (district, city, etc.) of migrants belonging to a particular social group (for example, protests against the construction of a mosque in an «orthodox city»);
    • · O. quoting explicitly xenophobic discourse and texts without comments to determine the division between the interviewee’s and journalist’s position; similarly — providing newspaper space for explicitly nationalist or other propaganda without editorial comment or further debate.
    • P. charges of an attempt to power seizure or territorial expansion (literally, as opposed to an appeal to prevent consolidation in the region);
    • Q. citizenship denial (e.g., referring to some Belarusian citizens as foreigners depending on their ethnic identity)
Types of HS (the scale by “Xenomonitor” initiative)
  • ·implicit or explicit calls for violence, violence classified as a valid means of solving problems; 
  • ·mentioning a distinguished or attributed sign of a group/community in reports of  criminal nature;
  • · statements or arguments about the need to provide special status;
  • · mentioning a distinguished or attributed sign of a group/community in reports on mass violence acts;
  • · charges of an attempt of power seizure, territorial expansion, domination, an indication of the negative impact of «them»;
  • · statements or arguments about the need for proportional representation;
  • · conclusions about the quality or the actions of others on the basis of allocated or attributed signs;
  • · use of expressive and rough language 
  • ·attribution to a group or community of qualities or actions, extended in time and space; 
  • · comparing groups and communities;
  • ·use of derogatory language;
  • ·personification (attributing to groups and communities personality characteristics);
  • inappropriate mentioning and updating of the signs of belonging to a group / community;
  • substitution of the concept of «human rights» with the concept of «communities rights»;
  • specifying territorial connections
  • Unification and joining individuals in a group;
  • Highlighting the importance of differences between groups and communities;
  • A combination of several interconnected signs or stating the need for such a combination
Tolerance of the article
  • · appeals and positive arguments for physical violence and destruction;
  • ·  appeals and positive arguments for removal from common living space (eviction, prevention of arrival);
  • · refusal to recognize the differences, specific features («they» are the same as «us», such as «Ukrainians and Russians have the same cultural foundations»);
  • · Statements of acceptability and desirability of assimilation and unification («Let them not stand out»);  
  • ·calls to abandon a dialogue or discussion of its impossibility («there’s no use talking to them,» «they will never understand»);
  • ·appeals to and discussing in a positive way exclusion from the sphere of public space and decision-making or limitations on participation in public sphere and decision-making (deprive of the right to vote, ban to express your view in mass media);
  • ·presenting the position of one side only;
  • presenting two opposite or contradictory positions without an attempt  to compare or analyze them, without seeking common grounds;
  • ·depicting the positions as mutually exclusive, suggesting “either-or” choice, without trying to build up a third version, combining the existing ones or alternative to them