Polish martial law of 1981 as seen on Facebook. A comparative analysis between the Facebook posts of political parties

ABSTRACT

This research focused on better understanding how social media are used in politics in the context of dividing historical event such as the Martial Law in Poland in 1981. A mixed method approach was used, including content analysis and logistic regression. The results showed that while there is a difference in which parties post about Martial Law in social media, the content and the sentiment of the discussion is quite similar. Readers' interaction was connected with the content's publication time. This study showed how political debate differs between social media and traditional media while presenting the antecedents of users' interaction.

KEYWORDS:

• Polish martial law of 1981
• political debate
• social media
• Facebook

Introduction

Within the context of contemporary political debate, the introduction of Martial Law in 1981 remains one of the most discussed historical events (Krupa 2021a). The actions undertaken by the communist government in order to secure the diminishing influence of the Polish United Workers Party and to quell the raising power of the Solidarity Trade Union was to end with widespread imprisonments, arrests and, in several cases, civilian deaths (Sowa 2011). As many democratic governments that came to power in the 1990s did little to address the calls for justice, the problem of the 1981 Martial Law still remains a burning issue for political discourse, one only recently surpassed by the 2010 Smolensk Disaster. The anniversaries of the introduction of martial law are a trigger for political manifestations, and the phrase “martial law” functions as a rhetorical figure used in disputes completely unrelated to the history of the Polish People's Republic. Only in the last few years, comparisons to the Martial Law of 1981 made by Polish parliamentarians referred to: the conflict around the Constitutional Tribunal, the management of Public Television, restrictions introduced by the government related to COVID-19, the Ombudsman's assessment, and even the situation of wild boar in Polish forests. Despite the appearance of new documents (both domestic and foreign), there is still a debate within the Polish press as to whether the actions undertaken by the Communist government were necessary given the possible doom of a Russian invasion, or rather were they merely a deliberate campaign aimed at Polish citizens (Krupa 2021b). As the debate perspectives align with the political profile of press publishers, they can also be seen in political debate happening in other medias or in the Polish parliament. At the same time, with each year there is a change in citizens’ perception of the mentioned events – in general falling more on the side that Martial Law in 1981 was an illegal decision (Pankowski 2021).

As social media are becoming more popular with users of all generations, more and more political discourse is on the Internet. A shift to social media usage as the main venue for gathering information and opinions in the context of Polish citizens has been noted and investigated by Stachowiak-Krzyżan (2019). At the same time, Farkas and Bene (2021) have investigated the strategies politicians use in their online communication. A similar approach in analysing political debate in online venues was also undertaken by Marquart, Ohme, and Möller (2020), Carlisle and Patton (2013), Kalsnes (2016) or Herring and Dainas (2017). In a similar fashion, Werner (2019) went on to establish how historical knowledge impacts the ways Polish citizens use social medias. It was found that Poles are usually looking for content connected with emotional engagement and for content connected with patriotic identity. However, despite the broad subject coverage, there are still certain research gaps to be filled. M. A. Xenos, Macafee, and Pole (2017) have urged for a more detailed analysis of political posts content stressing the need for a more qualitative focus. Alashri et al. (2016) have suggested that more research into political issues and the sentiment connected to them in social media should be done on websites other than Twitter. Carlisle and Patton (2013) have called for a broader investigation into the ways social media users are interacting with each other’s content in the case of political posts. In a similar fashion, Ben-David and Soffer (2019) have pointed out that further research should focus more on relationships explaining how users are interreacting with each others’ comments in social media.

Therefore, this research focused on better understanding how social media are used by politicians and commentators alike in the context of dividing historical events. The study’s goal was firstly to identify what type of content is posted by political parties during anniversaries of dividing historical events. Secondly, the research aimed to establish how readers are commenting and how they are interacting with other users’ comments under the mentioned posts. To achieve this aim, a mixed method study was conducted with the use of data acquired through web-scrapping of content posted by the main Polish political parties on the 39th and 40th anniversary of 1981 Martial Law. However, it should be emphasised that the main aim of the article is not to compare the debate in traditional and social media The study used content analysis, logistic regression and sentiment analysis to verify how different aspects influence the ways Polish citizen generate and react to political content related to important historical events.

This study makes a significant contribution to the existing literature by establishing how political debate in social media differs from traditional (press and television) media debate in the context of 1981 Martial Law. The given research also provides an insight into the user generated content created by social media readers by analysing the key themes within them and comparing them across different political parties. Finally, the current study provides further evidence of antecedents making users’ comments more engageable with other readers. In order to make such a contributions, the objectives of the study were as follows:

1. To critically compare which political parties are creating content related to 1981 Martial Law and how they present this event

2. To verify if there is a difference between the themes appearing in comments under each political party’s posts

3. To establish which elements of users’ comments influence readers to engage with them

Subject literature review

Polish martial law of 1981

In Poland, in 1980, “Solidarity” was founded – a movement gathering almost 10 million Poles demanding not only the improvement of living and social conditions, but also the creation of a free trade union. The communist authorities could not afford the functioning of a gigantic social movement beyond its control. After the change of the First Secretary of the Polish United Workers’ Party (Stanisław Kania in place of Edward Gierek), the communist authorities considered using force to restore the monopoly of power in the state, but the new leader opted for breaking up “Solidarity” and introducing agents to it. However, this strategy did not bring the expected results, so in 1981 a decision was made to use force. General Jaruzelski – Minister of National Defense, who in 1981 became Prime Minister and First Secretary (in place of Stanisław Kania), decided on such a radical solution.

The first sign of the introduction of Martial Law, was that telephone communication was blocked on December 12th, 1981, and from around 11.30 pm (before the formal declaration of Martial Law itself). The decree on Martial Law signed by the Polish Council of State introduced many restrictions: curfew, censorship of correspondence, ban on press publishment, militarisation of workplaces, suspension of trade union and associative activities and a ban on gatherings or demonstrations. The legality of the decree on Martial Law raises doubts, as the document is dated December 12th (1981), while the State Council only accepted the decree on December 13th. Firstly, the act was announced in the Journal of Laws of the Republic of Poland of December 17th (1981), and yet was applied on December 12th. This proves that citizens did not have an opportunity to familiarise themselves with the new law, which had already come into force. Secondly, it is problematic to determine whether the State Council had the right to impose Martial Law. It is worth recalling that the Council of State could issue decrees only between sessions of the Sejm (Polish Legislative Body), and its current session was not closed (Sowa 2011).

Governmental power was taken over by the Military Council for National Salvation (WRON), which had no legal regulations in the Constitution of the Polish People's Republic – it consisted of 22 officers of the Polish People's Army. The influence of the military forces on the political life of the country had significantly increased (Eisler 2018) and “ad hoc” court proceedings were introduced. In addition to internments, there were also arrests . At the same time, during the interrogations, there were reports of incidents involving blackmail, threats, and beatings – some of which ended up being fatal (Paczkowski 2006). The trials were often propaganda in their nature, with the most severe sentences ending with the death penalty in absentia. Martial law also included the pacifications of striking factories and mines, such as “Manifest Lipcowy” (4 miners shot) or “Wujek” (9 killed and 49 wounded miners, 41 policemen were wounded).

Martial law was abolished on July 22 (1983). Over its duration, 10 132 people were interned. Between December 13th (1981) and August (1982), 55.8 thousand people were fired from their workplaces (Paczkowski 2006). The number of victims of Martial Law is difficult to determine, because one should remember not only about those shot, beaten to death or taken care of by “unknown perpetrators”. There were also people who lost their lives due to telephone communication being cut off, as ambulances could not reach them (Sowa 2011). In 2011 – 30 years after the initial introduction of Martial Law – the Constitutional Tribunal fully recognised the decrees introducing Martial Law as incompatible with the Constitution of the People's Republic of Poland.

The issues of dispute in the debate on Martial Law include, first of all, the reasons for its introduction as well as the assessment of both the Martial Law itself and the people implementing it. In 2021, Polish society was greatly divided in the assessment of the events that occurred in 1981-1983: 33% assessed General Jaruzelski's decision as right, 37% as wrong. Interestingly, back in 1994 54% considered W. Jaruzelski's to be the “right” decision and 23% as “wrong”. However, it should be noted that in 2021 only 47% of respondents were able to indicate the date of the introduction of Martial Law in Poland (Pankowski 2021). Even though Polish society has such scant knowledge of the facts, the symbolism of Martial Law is often used in public discourse to compare the current political situation with that of the 1980s (Krupa 2021a).

Martial law in the mass media

The mass media not only mediate the transmission of information – as they play the role of the primary broadcaster. They also become an independent, separate political actor: with their own messages, values and accepted views (Street 2006). They create their own reality, which may not have much in common with that experienced empirically (Schulz 2006), and is of fundamental importance for individuals and societies. The media shape collective images by creating a symbolic reality – the world is explained and interpreted within a certain conceptual scheme (Skarżyński 2002). The symbolic world allows us to understand lived experiences and creates a network of fixed points of reference. Symbols shape the habit of unreflective acceptance and are used to manipulate the empirical sense of events – the symbolic aspect becomes the real sense and is perceived as real (Churska 2005). The symbolic universes created by the media and the stories told by the media that indicate the heroes and the guilty are of fundamental importance – collective memory is not based on an individual experience of a situation, but on a message that we believe and take as part of our identity (Szalkiewicz and Sokołowski 2019). Symbolic universes differ in their perception of actions undertaken by different actors as well as in understanding their motivations.

In the case of Martial Law, at least five symbolic universes (SU) can be distinguished, which were created or maintained by Polish national papers. Lets introduce the mentioned media groups. First – “Trybuna Ludu” (later: “Trybuna, and “Dziennik Trybuna”), was an organ of the Central Committee of the Polish United Workers’ Party. This newspaper has always been associated with the left, but its message and form of support differed depending on who it was led by. Some editors supported democratic changes, undertook a balanced criticism of privatisation and the presence of religion in public life, and some used harsh rhetoric and an aggressive tone (Mielczarek 2007).

The second group was the establishment of “Gazeta Wyborcza”. Its creation was one of the results of the talks held in 1989, when the Solidarity opposition was permitted one daily and one weekly newspapers. After the presidential elections of 1990 and the early parliamentary elections in 1991, Gazeta Wyborcza began to be perceived as an organ of the Democratic Union Party, later transformed into the Freedom Union. In the presidential election of 1990, it supported Tadeusz Mazowiecki (first non-communist Polish prime minister in III Republic of Poland), in 1995 – Jacek Kuroń (minister of Labour and Social Policy, co-founder of the Workers’ Defence Committee in 1976), in 2000 its position towards Aleksander Kwasniewski (co-founder of the Social Democracy of the Republic of Poland), who was seeking re-election, can be considered neutral, and in 2005 it favoured Donald Tusk (co-founder of the Civic Platform). In relation to the post-communists, i.e. mainly the Social Democracy of the Republic of Poland, the newspaper had a more conciliatory attitude – it did not call for decommunisation, lustration, political purges or being made accountable for the past. (Mielczarek 2007) In presidential election 2015 it showed unequivocal support for Bronisław Komorowski, and 2020 – Rafał Trzaskowski (Krupa 2021b).

The third group was the establishment of “Rzeczpospolita”. It has been published since 1982. After the transformation of the political system (1989), it created a place for presenting ideologically different views. The ideological profile of the newspaper depended to some extent on its editor-in-chief. Articles in “Rzeczpospolita” often gave rise to nationwide public debates, e.g. on the future of the EU (Kamińska 2013). The fourth group is headlined by “Nasz Dziennik” which is associated with “Radio Maryja” and the circles of Father Tadeusz Rydzyk. The daily's programme line can be described as patriotic-Catholic-conservative, anti-EU (Mielczarek 2007), promoting the decommunisation of Polish political life and settling accounts with the communist past (Krupa 2021c). “Nasz Dziennik” once supported the League of Polish Families. In 2015 and 2020, in the presidential elections, the daily supported Andrzej Duda (Law and Justice) (Krupa 2021b).

In the first “SU”, one created before 1989 by the “Trybuna Ludu”, Martial Law was a salutary operation that prevented civil war and unheated homes in winter. General Jaruzelski and his associates are presented as heroes. Later on, in the second “SU”, which was represented after 1989 by “Trybuna”, “Dziennik Trybuna”, Martial Law was a tragic event for both sides and the blame for its introduction rests with both the authorities and the opposition. General Jaruzelski took a great burden on his shoulders, he was the statesman and patriot who saved Poland from Soviet invasion; he deserves gratitude and respect. Martial Law is often juxtaposed with the May Coup (1926; it is emphasised that the latter lasted much shorter and resulted in many more victims), the suppression of the Hungarian Uprising (1956) and the Prague Spring (1968). It is emphasised that The Polish People's Army could resist in the event of an external intervention by Warsaw Pact troops. The figure of Martial Law is also used to describe the socio-political reality under the rule of Law and Justice after the double election victory of 2015.

In the third “SU”, created since 1989 by “Gazeta Wyborcza”, Martial Law is an event involving a lot of harm and damage, but the people responsible for its introduction a few years later decided to talk at the Round Table, which was a “blessing” for Poland. In Gazeta Wyborcza, Martial Law is associated with events that refer not only to the socio-political conflicts taking place in the Polish People's Republic, but also to the Second Polish Republic, and even the revolutions in England and France (17th and 18th centuries). The issue of Martial Law is also mentioned on the occasion of such topics as: decommunisation, lustration, and systemic transformation (1989). In the fourth “SU”, martial law was a method of defending the communist power in Poland. General Jaruzelski demanded military aid from the USSR. The dominant position in Rzeczpospolita was that Poland was not threatened by the intervention of the Warsaw Pact troops. Jaruzelski was a complex figure, but he lacked courage and shared responsibility for many crimes because the introduction of martial law was his personal choice. In “Rzeczpospolita” also representatives of the former communist authorities spoke, hence there were also opinions favourable to Jaruzelski and justifying his decision due to a certain Soviet intervention. Who was the editor-in-chief depended on how often the subject of martial law was discussed. Martial Law also appears in the context of lustration, justice, journalists, and the Round Table (1989).

In the fifth “SU”, represented by “Nasz Dziennik” and here since 1998, Martial Law is a clear evil committed upon the Polish nation by the “traitors” of the motherland who have never been held responsible for the harm done, with the number of its victims large. General Jaruzelski is called a dictator. In “Nasz Dziennik”, Martial Law is repeatedly combined with the Round Table (1989) and the process of political transformation. There are also references to the situation of Poland in the eighteenth century (The Targowica Confederation), World War II (Katyn) and Czechoslovakia (the Soviet intervention in 1968). Martial Law also appears in the context of philosophical deliberations on the nature of man, with the problem of the burial of the dead or abortion. (Krupa 2021a) (Table 1).

Table 1. Symbolic universes created by national dailies.

Perspective on:	Trybuna Ludu (I) / Trybuna (II) / Dziennik Trybuna (II)	Gazeta Wyborcza (III)	Rzeczpospolita (IV)	Nasz Dziennik (V)
Martial law introduction	Affirmative	ambivalent and critical	critical and ambivalent	critical
Wojciech Jaruzelski	Hero who saved Poland	He deserves gratitude	He had a choice, but he lacked the courage	traitor, criminal
events that can be compared to the 1981 Martial Law	May Coup 1926, Hungarian Uprising 1956, Prague Spring 1968	The Round Table 1989, May Coup 1926, violations of the law in the Second Polish Republic, social protests in the times of the Polish People's Republic	The Round Table 1989, Prague Spring 1968, lustration, catholic church, martial law judges	The Round Table 1989, the Targowica Confederation 1792, the Katyn massacre 1940, victims and persecution in the times of the Polish People's Republic

For the purposes of the article, it was assumed that the main opposition party, i.e. Civic Platform, is closest to the Symbolic Universe III (Gazeta Wyborcza). At the same time, the Law and Justice party, i.e. the main ruling party, is closest to the V universe. Universe IV is ideologically located between SU III and SU V (representing positions less radical than SU V, but more critical towards martial law than SU III). This due to the fact that Rzeczpospolita has been trying for many years to develop a certain independence and autonomy (Milczarczyk 2019). This stands in opposition to Gazeta Wyborcza or Nasz Dziennik, which openly support specific political formations (Krupa 2021b). Because of the fact that it cannot be easily assigned to one of the parties in the political dispute, SU IV created by Rzeczpospolita will not be as a major subject as SU III and V.

Social media in political debate

As social media are becoming an ever increasing part of people’s day-to-day lives, they have also started to influence more their actions, decisions, and ways of communicating. In the context of Polish society, the influence of social media has been the subject of research by Stachowiak-Krzyżan (2019) who established that 44% of social media users decided to buy a product they hadn’t planned to as a result of adverts they saw on social media. Similarly, 76% of Polish Internet users declared that the positive feedback about services or products voiced by their online friends significantly influence their purchasing decisions (Stachowiak-Krzyżan 2019). But the power of social media is seen not only for commercial purposes, but also to be used by politicians and political parties. As such, the use of social media in political discourse has already been the subject of many studies.

Farkas and Bene (2021) have analysed the image-based communication of Hungarian political parties during the 2018 election campaign and showed how Facebook is used to personalise communication with followers. Marquart, Ohme, and Möller (2020) have identified that young Danes who are following political actors on social media are more inclined to engage in political campaigns, especially if their friends are also active within similar topics. On the other hand Kruse, Norris, and Flinchum (2018) have established that the general public as a whole is less inclined to share their political views and engage in political discussions, mostly due to fear of online harassment and the view that social media is a platform for less serious political discussion. What is more, the study showed that when facing people of views different than theirs, both Millenials and Generation X social media users were inclined to unfollow or unfriend them to avoid facing difficult discussions (Kruse, Norris, and Flinchum 2018).

The question, however, arises – what are the factors influencing political engagement’s frequency and ways on social media sites? Rainie et al. (2012) have showed that people of more moderate views are less inclined to participate in political debate, while citizens of more left- or right-wing perspectives more often decide to engage in online discussions. Carlisle and Patton (2013) went on to establish that size and the importance of the elections directly influence the level of engagement on social media. What is more, gender and income of parents (usually the main indicators of political engagement) was showed to not be of statistical significance in the online environment (Carlisle and Patton 2013). Xenos, Vromen, and Loader (2014) have also investigated the sociodemographic antecedents of online political engagement and found significant differences in factors such as civic education at home and the general social media use of citizens. However, all of the aforementioned studies were based on surveys and focused solely on engagement antecedents related to the personality and background of people. Skoric et al. (2016) advocate further research into the subject, which would go beyond the typical approach of surveying college students. On a similar note, Carlisle and Patton (2013) have also called for more generalised samples to be used in further research on the subject. At the same time Kalsnes (2016) has established that when it comes to politicians communicating on social media, their strategies vary depending on the perceived media attention, online reputation risk and access to resources. Therefore, it is hypothesised that:

H1: The social media approach to 1981 Martial Law of Polish political parties (representing views of different SU) will differ in terms of their post frequency and content

Furthermore, Adamik-Szysiak (2019) has identified, when analysing Polish politicians activities in social medias, that within the context of the Polish political world, social medias are used in an “attempt to impose a specific vision of social and political reality around potential voters”. It would suggest that different political parties will have a different approach in how they present the events of 1981. Therefore, it is further hypothesised that:

H1a: The events of Martial Law of 1981 would be brough up in social media posts mainly by parties that are critical towards it

When it comes to the subject of social media communication used by political parties, Boukala’s study of 2014 identified that parties’ political identities can serve as a prism through which a divided past can be seen. Subsequently, this has influenced the content political parties have posted in the Internet. In this context, it is hypothesised that:

H1b: The social media posts connected to the Martial Law of 1981 posted by all parties would all be of similar tone and message

When it comes to the subject of the content of the comments and the language used, some researchers have already undertaken investigations. The issue of opinion mining in the context of political election has been researched by Karami, Bennett, and He (2018) who used word frequency analysis in order to establish key topics discussed in social media. Also Ben-David and Soffer (2019) went on to identify that there were differences in comments’ topics across different platforms.

H2: Social media content related to Martial Law of 1981 posted by different parties will observe different commenting patterns

Posts of political parties on Facebook are commented on by both supporters and opposers of the party (Bustikova et al. 2020). In case of Poland, according to research by Stetka, Surowiec, and Mazák (2019), 68% of women's comments concerning the party that authored the post were positive, 14% ambivalent and 18% negative. In the case of men, these were 52% positive, 15% ambivalent and 33% negative. However, if the comments concerned other parties, in the case of women they were 7% positive, 24% ambivalent and 69% negative, and in the case of men: 10% positive, 36% ambivalent and 54% negative. In turn, research by Kim et al. (2021) focusing on comments under the posts of news outlets in the USA showed that more toxic (hateful, more polarising or antagonising) comments were observed under posts representing a liberal worldview, and fewer under conservative ones.

The fact that certain words are used in different frequencies by different political players has also been identified by Lin and Kant (2021) – here in the area of social media connected to citizen participation. Peng (2021) has also suggested more in-depth studies employing computational textual analysis in the context of political content on social media. In this context, it is hypothesised that:

H2a: Comments under the posts of opposition party will be focusing more on the past and remembrance

H2b: Comments under the posts of governing party will be focusing more on the contemporary political topics

While many researches has focused on the content of social media posts, there is also an academic debate into how readers engage with this type of content. When it comes to social media engagement in the political context, M. A. Xenos, Macafee, and Pole (2017) have investigated how the tone (measures either posts of a promotional, attacking or contrasting nature) and periodicity (measured in the frequency of new post appearing) influences participation in online debates. The research showed that sentiment presented in the posts resulted in different scales of responses. At the same time, the longer a post was visible, the more attention it attracted over time. Interestingly, posts containing media files such as images or gifs usually generated lower interest than other posts by the same author (M. A. Xenos, Macafee, and Pole 2017). Subsequently, the authors called for further research into user choices as to which content they respond to and in what way. Finally, Segesten et al. (2020), who have focused their studies on the post and comments valence, identified that disagreement in the comment feed increases readers attention but at the same time decreases the likelihood of the post being shared by the reader. Therefore, it is hypothesised that:

H3: The characteristics of the comments relating to 1981 Martial Law-related posts influence readers engagement

Firstly, the issue of comment age was taken into consideration. In the context of tourism-related social media, Akbarabadi and Hosseini (2020) have shown that more people interact (by choosing a review as helpful) with the least oldest and the newest opinions. On the other hand, M. A. Xenos, Macafee, and Pole (2017) have found that more reactions were acquired by content that was visible for a longer time. As such, it is hypothesised that:

H3a: The time of comment submission influences readers engagement

Furthermore, Ben-David and Soffer (2019) when analysing comments under news articles found differences in their length depending on the venue they were written on. As such, they possessed different amounts of detail and content. Similarly, Filieri, Alguezaui, and McLeay (2015) have established that user generated content posted online was considered more trustworthy the more details and information it contained. Because of these, it is hypothesised that:

H3b: The length of the comment influences readers engagement

The issue of using “graphicons” (GIFs, emojis, graphics or icons) in social media comments was the subject of a study by Herring and Dainas (2017). It was identified that they were mostly used for conveying reaction sentiment or in changing the tone of the conversation. However, the researchers have suggested investigating the use of “graphicons” in other social contexts. As such, it is hypothesised that

H3c: The use of the “graphicons” influences readers engagement

As was previously mentioned, the comments valence was another possible antecedent of user interaction. In the context of so called hard news topics (e.g. politics) readers attention decreased if the comments showed disagreement (Segesten et al. 2020). At the same time, Alashri et al. (2016) have focused on comments sentiment during the 2016 elections in the United States of America. It was identified that comments sentiment differed in comments feeds under posts authored by different politicians (Republican/Democratic). Therefore, it is hypothesised that:

H3d: The sentiment of the comment influences reader engagement

Moreover, the issue of user engagement in the online discourse has been raised by Platonov and Svetlov (2021). The aspect of content readability was analysed in how it influenced readers engagement. However, it turned out to be a poor predictor within the general construct. On the other hand, Liang, Schuckert, and Law (2019) have shown that social media posts in the field of tourism are perceived as more helpful if they are more readable. Because of these, it is hypothesised that:

H3e: The comment readability influences the reader engagement

Methodology

Research design

In order to address the research question of evaluating the political discourse within the context of 1981 Martial Law 10 hypothesis were tested. In that sense the research design adopted a mixed method approach and the research type was explanatory. Such an approach had previously been successfully used in the field of political sciences by Lin and Kant (2021). Given the rather wide context of the analysed problem, a mixed approach was chosen to better understand the ways political parties related to the issue of the 1981 Martial Law (qualitative approach) and the ways in which Polish citizens were commenting on the posted materials (quantitative approach). The developed research procedure followed three stages.

Firstly, a content analysis of Polish political parties’ posts published on Facebook on 13th December 2020 and 13th December 2021 was undertaken. An empirical study approach was used in this part of the study, here due to the limited sample volume and the rather short nature of the content posted on social media (Idzik and Klepka 2019). A similar approach had been previously used by Weltevrede, Helmond, and Gerlitz (2014) in the context of political studies in the online environment. Each of the encountered post was archived by downloading the text along with any accompanying photos published by the authors. Subsequently, the content gathered was compared across the different political parties and different timelines. In the analysis, the thematic subjects identified during the subject literature review stage were used to better understand the sentiment carried by each of the posts (Idzik and Klepka 2019).

Secondly, the content of the comments was investigated through the identification of the most frequently used keywords in comments under each of the posts in the analysis. Such a frequency analysis approach has already been undertaken in social media research by, for example, Lin and Kant (2021) or H. Wang et al. (2012). The gathered data were analysed with the use of the MaxQData application, with a separate analysis undertaken for each of the posts. Most common Polish words were excluded from the analysis based on the available word data banks. In a similar fashion, a lemmatisation procedure was undertaken to group words of similar meaning appearing in different conjunctions. Finally, the ten most frequent words used in each of the comments block was exported for further analysis.

Thirdly, a quantitative analysis was undertaken to evaluate the antecedents of the popularity of comments posted under each of the archived posts. A web-scrapping tool called “Export Comments” was used to download the data from each of the Facebook posts. As Speckmann (2021) has shown, web-scrapping is an inexpensive method of acquitting large datasets that can be used for testing an existing hypothesis in social sciences fields. The process happened 2 weeks after the publishing of each post, as it allowed for a proper saturation of comments in each of the cases. With the use of the “Export Comments” application the following information for each of the comments was acquired (Table 2):

Table 2. Data extracted through web-scrapping.

Name	Description
Comment content	The textual and graphical content of the comment
Comment author	The name and link to the profile of the comment’s author
Post ID	The ID number allowed for classification as to whether the comment was posted under the original post or under a different comment
Reactions number	The number of reactions the comment gathered from other readers
Time since publication	The time between the publishing of the original post and the comment

In order to analyse the determinant of the comment popularity, logistic regression analysis was used. Logit regression is a statistical method used to analyse the relationship between a binary response variable (the comment receiving or not receing reactions) and one or more predictor variables. The results of the logit regression analysis are usually presented in the form of coefficients. Each coefficient represents the amount by which the log-odds of the outcome variable change for a one-unit increase in the corresponding predictor variable, holding all other predictors constant. A positive coefficient indicates that as the predictor variable increases, the likelihood of the outcome variable occurring also increases. Moreover, in order to better understand the influence of the predictor variables, Average Marginal Effects (AME) were computed for each (with all other covariates held at their actual values). In this approach the research was inspired by the works of Wang and Sant (2022) in analysing big data obtained from social media and concerning social issues and by work of Norton, Dowd, and Maciejewski (2019) in most common way of reporting marginal effects in the studies. Based on the hypotheses development stage, the following variables were used in the regression analysis (Table 3).

Table 3. Variables in logistic regression analysis.

Name	Measurement	Variable type
Comment’s popularity	Had a comment received reactions (1 – Yes/0 – No)	Dichotomous
Comment’s age	Time between the publication of the post and the publication of the comment	Ordinal
Comment’s length	Length of the comment in characters	Discrete
Comment’s type	1 – Main comments / 0 – Subcomment	Dichotomous
Comment’s media use	1 – comment used graphicons/ 0 – comment didn’t use graphicons	Dichotomous
Comment’s valence	The sentiment measured as negative (−1)/neutral (0)/positive (1)	Ordinal
Comment’s readability	Readability of the comment measured on the FOG scale	Discrete

The data about comments’ popularity, age (how long it was visible) and media use was derived from the initial data gathering stage. The comments’ type reffered to fact whether the comment was published directly under the main post (1 – Main comment) or as a reply to already existing comment (- - Subcomment). The comments’ length (measured in the number of signs) was calculated using Excel functions. For the comments’ readability score, a FOG scale evaluation was conducted with the use of the Jasnopis application as previously used by Jezierski (2021). The FOG scale measures how easily the text can be understand by assigning score between 6 (text understandable by 6th garder) to 17 (text understandable to college graduate) Comments valence was coded manually, as the data set acquired was in Polish and the existing tools for sentiment classification either required longer texts that those existing in comments or are designed for sentiment analysis in English. Four types of sentiment were graded: general sentiment, sentiment towards ruling party, towards main opposition party, towards Martial Law of 1981. The regression analysis was conducted with the use of the SPSS application and the marginal effects were estimated with the R Studio package MFX.

Sample

The topic of 1981 Martial Law was chosen for the analysis as it represents a very divisive issue that is still a theme of contemporary debate within Polish society (Krupa 2021a). Parties from different political wings present different perspectives on this issue and their social media content is a reflection of the mentioned division. At the same time, during the analysed years of 2020 and 2021, Polish society’s political views were significantly divided between the main governmental party (Law and Justice) and the main opposition party (Civic Platform). Consequently political discourse in traditional and social media was increasingly radicalised making for an important topic to be academically researched. Facebook posts and profiles were chosen as the subject of the analysis, as Facebook remains the most popular social media portal in Polish society. In 2021 it was ranked as the third most often visited website on the Polish Internet, preceded only by the Google Search Engine and Youtube. By choosing the most popular social media portal, it allowed for a better generalisation of the final results.

The web-scrapping procedure resulted in the download of 7 posts from 2020 and 12 posts from 2021. Subsequently, four posts were chosen for the comments analysis – two from the main governmental party and two from the main opposition party. Such an approach was chosen because the polarisation of political views in the online environment is becoming an even more important topic in current academic investigations (Modgil et al. 2021; Salahudin et al. 2020). For the comments web-scrapping results 1134 comments were collected in 2020 and 842 comments were collected in 2021.

Results

The posts of Law and Justice in 2020 concerned the commemoration of the victims of Martial Law, and on the website you could find photos with a description of the most important party politicians (Jarosław Kaczyński, Mateusz Morawiecki) who on that day were at the graves of those murdered and lighting a candle in the window. In 2021, Law and Justice posts regarding Martial Law were also focused on the victims, and party politicians laid flowers on their graves. Posts that were not directly related to the anniversary of the imposition of Martial Law include: information about the investment programme under the “Polish Deal”, information about the meeting of Prime Minister Mateusz Morawiecki with the German Chancellor Olaf Scholz and his meeting with the V4 group, as well as laying flowers at the monument to Lech Kaczyński (the President of the Republic of Poland in 2005–2010, the brother of Jarosław Kaczyński).

The post of the Civic Platform in 2020 mainly concerned the issues of the drastic restriction of civil liberties and the opposition of people who decided to fight for the freedom of subsequent generations. In 2021, the post contains a short message, “On the 40th anniversary of this dark event – in honour of the memory of the victims”. The other two are: an almost three-minute recording of the Civic Platform action plan and the change of the background photo to: (Law and Justice = High Prices; “Law and Justice = Drożyzna”.)

The Left's posts in 2020 and 2021 were very similar and concerned with the celebration of the finalisation of Poland's accession to the structure of the European Union on December 13th, 2002, carried out in Copenhagen by representatives of the Democratic Left Alliance- Labour Union-Polish People's Party government. The second post from 2021 is a more than half-hour meeting with a deputy and senator from the Left “#Kawusia Lewicy” (leftie coffee) regarding the protection of health and animals. Similarly, the post of the Polish People's Party in 2021 concerned the celebration of the death of a pre-war Polish People's Party activist – Stanisław Mikołajczyk. A tabulatory summery of the post can be seen in Table 4

H1a: The events of Martial Law of 1981 would be brough up in social media posts mainly by parties that are critical towards it

Post-Solidarity parties (Law and Justice and Civic Platform) in 2020 and 2021 placed information on the introduction of Martial Law on their profiles. The parties with a post-communist origin (Left, and Polish People's Party) did not include such information. (Polish People's Party post from 2020 could not be accessed). As such, Hypothesis 1a was supported. The analysis confirmed that parties whose heritage is connected to the regime in power during the events of 1981 try to omit the commemoration of it from their social media spaces.

Table 4. Political parties’ social media posts on the 13th of December.

–	Number of posts related to Martial Law (total number of posts on that day)
Year/party	Law and Justice	Civic Platform	Left	Polish People's Party
2020	4 (4)	1 (1)	0 (2)	no data
2021	3 (7)	1 (3)	0 (1)	0 (1)

H1b: The social media posts connected to the Martial Law of 1981 posted by all parties would all be of similar tone and message

If the parties decided to publish posts on Martial Law, there are no major differences in their content. An interesting example, however, is the Left, which on December 13th commemorates a completely different event, both in 2020 and 2021: the finalisation of Poland's accession to the structure of the European Union, which took place on December 13th 2002, carried out in Copenhagen by representatives of the Democratic Left Alliance- Labour Union-Polish People's Party government. Thereby, Hypothesis 1b was confirmed. This result implies that parties which decide to commemorate the events of 1981 are looking at them from the same perspective – the post-solidarity perspective.

H1: The social media approach to 1981 Martial Law of Polish political parties (representing views of different SU) will differ in terms of their post frequency and content

The communication of political parties on Facebook regarding Martial Law differs – some parties have not decided to commemorate this event at all (Polish People's Party, Left), and the Left celebrates a completely different event on the same day. Even though Law and Justice and the Civic Platform publish similar posts regarding Martial Law, Law and Justice does it in larger numbers and focuses mainly on commemorating the victims – while the Civic Platform emphasises civil liberties. One could risk the claim that for Law and Justice Martial Law has a certain martyrdom dimension. As a result, Hypothesis 1 was deemed supported.

When it comes to the frequency analysis, the results are shown in Table 5.

H2a: Comments under the posts of opposition party will be focusing more on the past and remembrance

Words most frequently used in comments to opposition party were mostly focused on remembrance of the historical event. Martial was the 2^nd most used word, followed by law (4th), Jaruzelski (8th) and remember (9th). What is interesting is the context of comments under the opposition post is rather a low count of words connected to contemporary political debate. In the first ten most comments words there is no motion of either of the two main parties nor of any contemporary politicians names. Therefore, Hypothesis 2a was deemed supported.

Table 5. Frequency analysis result.

Comments to opposition party posts						Comments to government party posts
Word in Polish	Word in English	Total frequency	Total frequency %	No of comments with word	% of comments with word	Word in Polish	Word in English	Total frequency	Total frequency %	No of comments with word	% of comments with word
teraz	now	65	0.79	60	8.8	rok	year	119	0.67	81	6.27
wojenny	martial	55	0.67	41	6.01	polska	Poland	105	0.59	79	6.11
rok	year	47	0.57	43	6.3	PiS	Government party name (Law and Justice)	91	0.51	77	5.96
stan	law	46	0.56	35	5.13	pamiętać	remember	90	0.51	70	5.42
człowiek	man	42	0.51	39	5.72	wojenny	martial	81	0.46	73	5.65
Polska	Poland	38	0.46	31	4.55	stan	law	80	0.45	71	5.50
czas	time	35	0.43	29	4.25	Kaczyński	Kaczyński	72	0.41	66	5.11
Jaruzelski	Jaruzelski	34	0.42	15	2.2	człowiek	man	66	0.37	59	4.57
pamiętać	remember	33	0.4	31	4.55	inny	other	65	0.37	58	4.49
woman	kobieta	31	0.38	31	4.55	czas	time	59	0.33	47	3.64

H2b: Comments under the posts of governing party will be focusing more on the contemporary political topics

Mentions to the contemporary political debate are, however, prevalent in comments to the government party’s posts. Here the name of the ruling party is the 3rd most mentioned word and the surname of the current head of the ruling party is the 7th most used word. Interestingly, also the world Poland is more prevalent in the comments to government’s posts. As such, Hypothesis 2b was deemed supported.

H2: Social media content related to Martial Law of 1981 posted by different parties will observe different commenting patterns

The findings showed that there are substantial thematic differences in the comments content for opposition and government party posts. Comments to the opposition party were more connected with the idea of remembrance, time passing and the sole topic of 1981 Martial Law. On the other hand, the comments under the ruling party’s posts were more connected to the current political debate and more enrooted in the ongoing political conflict (especially in the context of the person of Jarosław Kaczyński). Thereby, Hypothesis 2 was supported. It seems that the way of engaging with political content that has historical context differs among social media users.

In the context of the quantitative analysis, the descriptive results are presented in Table 6. Comments from 2020 received more reactions both in the case of government and opposition parties. Interestingly, however, when both years are looked at, the average number of reactions is highly similar, with a statistical comment receiving about 2.31 reactions. When it comes to the pace of comment posting there is an inverse between 2020 and 2021. In 2020, the comments under the government party’s post were published more swiftly, while in 2021 the opposite was true. On average it took about 23 min after the post was published, for a statistical comment to appear. Looking at the length of the comments, in 2020 comments under the opposition party’s post were longer, but in 2021 they were significantly shorter. At the same time, in the case of comments under the opposition party’s post there was a sharp rise in the number of posts that employed “graphicons”. That trend is also true for the comments under the government party’s post. The analysis of the comments type shows that commentators publishing under opposition parties posts tend to engage in discussion with other commentators. On the other hand, people writing under the government party’s posts are more inclined to comment directly under the main post instead of commenting under other people opinions. The analysis of the sentiment recorded in the comments shows that mostly it was not to change between the years. The only exception is the case of sentiment to the events of 1981. In the case of both parties, the comments under the 2020s posts were more negative towards the Martial Law of 1981. Another important observation is the fact that negative sentiment towards both the government and opposition party dominated in all of the researched posts. Looking at the readability of the comments, the average score is around 6, which corresponds to a very understandable text, that can be easily comprehended by a 6th-grade student.

Table 6. Descriptive results.

Variable	2020	2021	Both years	2020	2021	Both years	2020	2021	Both years
Comment’s popularity [reactions]	2,72	1,74	2,31	2,67	1,47	2,31	2,94	1,98	2,31
Comment’s age [hours]	0,41	0,34	0,38	0,38	0,52	0,42	0,53	0,19	0,30
Comment’s length [signs]	143,04	126,27	135,89	137,04	161,72	144,51	166,36	95,53	119,59
Comment’s media use [0/1]	0,17	0,34	0,24	0,19	0,37	0,25	0,06	0,31	0,23
Comment's type [0/1]	0,39	0,51	0,44	0,47	0,64	0,52	0,06	0,40	0,29
Comment’s valance
Sentiment general	−0,58	−0,59	−0,58	−0,59	−0,57	−0,59	−0,50	−0,61	−0,57
Sentiment to opposition	−0,09	−0,09	−0,09	−0,08	−0,07	−0,08	−0,14	−0,11	−0,12
Sentiment to government	−0,37	−0,31	−0,34	−0,36	−0,40	−0,37	−0,41	−0,24	−0,30
Sentiment to 1981 events	−0,35	−0,22	−0,29	−0,28	−0,15	−0,25	−0,59	−0,27	−0,38
Comment’s readability	6,10	6,02	6,07	6,09	6,36	6,17	6,14	5,72	5,87

Focusing on the results of logistic regression, the assumption of a linear relationship between any continuous independent variable and the logit transformation of the dependent variable was verified to be positive with the use of the Box-Tedwell test. The Omnibus test showed the model to be statistically significant (χ2(9) = 354.06, p < .001). The explained variation in the dependent variable based on the evaluated model ranged from 16.6% to 22.2%. Although the results are not extremely high, this is to be expected in the field of the social sciences. What is more, the model correctly classified 68.5% of analysed cases, deeming it fit for further analysis (Table 7).

H3a: The time of comment submission influences the readers’ engagement

According to the analysis, the results time of a comment did significantly influence the engagement of the readers (p < .001) with each hour of the comment becoming older, the chance of the comment receiving a reaction decreased by a factor of 0.92. However, the average marginal effect showed that the decrease was only 2% with each hour. Nevertheless, Hypothesis 3a was deemed to be supported as posts such as ones analysed in the sample are usually easily observed throughout several hours. Based on the findings it seems that readers interact with the comments about the same time that the original post was written, with their engagement decreasing over time.

Table 7. Logistic regression results.

Variable		B	S.E.	Wald	df	Sig.	Exp(B)	95% C.I. for Exp(B)		Average marginal effect
								Lower	Upper
Step 1^a	Comment’s age	−0,085	0,008	113,622	1	0,000	0,918	0,904	0,933	−0,020
	Comment’s type	−0,666	0,113	34,621	1	0,000	0,514	0,412	0,641	−0,154
	Comment’s length	0,000	0,000	3,245	1	0,072	1,000	1,000	1,000	0,000
	Comment’s media use	−0,956	0,141	45,706	1	0,000	0,385	0,291	0,507	−0,206
	Comment’s valance general	0,136	0,088	2,396	1	0,122	1,146	0,964	1,362	0,032
	Comment’s valance to opposition	−0,130	0,147	0,785	1	0,376	0,878	0,659	1,171	−0,031
	Comment’s valance to govermant	−0,368	0,117	9,963	1	0,002	0,692	0,550	0,870	−0,087
	Comment’s valance to 1981 events	−0,276	0,110	6,344	1	0,012	0,758	0,612	0,941	−0,065
	Comment’s readability	0,007	0,008	0,728	1	0,393	1,007	0,991	1,024	0,002
	Constant	0,574	0,112	26,304	1	0,000	1,776

H3b: The length of the comment influences the readers’ engagement

The findings showed that the length (and because of it, the level of details included) of the comment do not have an influence on the readers’ engagement (p = .072). Consequently, Hypothesis 3b was rejected. This result implied that with comments being a very short and concise medium, longer comments do not usually convey more important or more engaging content. At the same time, this result should be contrasted with the use of the other media (GIFs, images, emoticons)

H3c: The use of “graphicons” influences the readers’ engagement

The findings also showed that the use of the “graphicons” significantly influences reader engagement (p < .001). However, this relationship turned out to be negative in nature – the use of GIFs or images resulted in a marked decrease in the chances for the comment to be reacted to with a factor of 0.385. Moreover, the marginal effect showed that the decrease was about −20%. Thus, Hypothesis 3c was supported. Such a result would suggest that big media files do not enhance the message of the comment and do not encourage people to interact with them.

H3d: The sentiment of the comment influences the engagement of the reader

According to the analysed results the comments’ overall sentiment as well as the sentiment towards the opposition party activities was not a significant predictor of the readers’ engagement with p equaling .122 and .378 respectively. On the other hand, the sentiment towards the government party actions as well as the sentiment towards the events of 1981 Martial Law were both shown to be statistically significant with p reaching .002 and .012 respectively. What is more, the marginal effects of these were –9% in case of sentiment towards governing party and –7% in case of sentiment towards the events of 1981 Martial Law. With such results, Hypothesis 3d was deemed supported. Comments that presented the events of 1981 in a positive way were less likely to generate user engagement with a factor of .758. In a similar way, comments presenting a negative perspective on the government party’s actions were more likely to receive a reaction from readers.

H3e: The readability of the comment influences the reader engagement

Based on the logistic regression results, the readability of the comment did not influence the reader engagement (p = .393). As such, Hypothesis 3e was rejected. This result could be explained by the fact that the readability of the comments in the sample was very similar across the whole sample and did not differentiate the reaction of the reader. Worth noting is the fact that most of the comments were very uncomplex in their nature.

H3: The characteristics of the comments to the 1981 Martial Law-related posts influence the readers engagement

In general, the logistic regression has showed that certain characteristics of the comments possessed a significant influence on the readers’ engagement. With the model being statistically significant and with Hypothesis 3a, 3c and 3d being supported, in overall Hypothesis 3 was supported by the analysis results.

Discussion

The attitude of political parties to the imposition of Martial Law in Poland, which decided on December 13th to publish a post on this subject on their Facebook profile, is no different. An element worth paying attention to is noting which parties omitted commemorating Martial Law on their social media profiles (the Left, Polish People's Party). Interestingly, there is no clear correlation with who honoured the memory of the victims of Martial Law and published a post on that day, and the perspective of parties’ voters on Martial Law’s introduction. According to the research, 42% of the Left voters, 40% of Civic Platform voters, 36% of Law and Justice voters and 34% of Polish People's Party voters consider General Jaruzelski’s decision to introduce Martial Law to be right. Considering it as wrong: 27% of the Left voters, 41% of the Civic Platform, 48% of the Law and Justice, and 28% of the Polish People’s Party. (Pankowski 2021). It can be stated that the parties that decided to post on Facebook on December 13th do not fit into the press narratives of Trybuna Ludu, Trybuna and Dziennik Trybuna. They are partially consistent with the narrative of Gazeta Wyborcza (violations of law, restrictions on civil liberties), Rzeczpospolita (victims) and Nasz Dziennik (focus on commemorating the victims and martyrdom). One should be cautious with comments made by political parties regarding Martial Law – it must be remembered that more than half of Poles do not know the exact date of its introduction. Moreover, the younger the Poles, the less knowledge they possess (Pankowski 2021). From the given research, it can be concluded that Martial Law in the symbolic universe of Polish participants in public discourse escapes from the reality of the 1980s and is localised in the present – it is used as a rhetorical figure to describe the current socio-political situation. This happens not only in public statements by politicians during a session of the Sejm interviews or posts on social media (Krupa 2021a), but is also used by “ordinary” Facebook users. The introduction of Martial Law in 1981 is thus compared, for example, with: (1) the tightening of the abortion law in Poland after the judgment of the Constitutional Tribunal on October 22th 2020, the reaction to it (women's strikes) and the brutality of the police in suppressing the demonstrations, (2) the lack of restrictions / restrictions related to the Covid-19 pandemic.

Simultaneously, posts published by political parties about the recent history of Poland arouse great emotions among Internet users. The language used in the comments is very aggressive, something confirmed by the research conducted by Zasada and Kamasa (2017). Urzędowska (2019) has also made a similar observation. In the cases analysed in this article, it is not only about vulgar epithets, comparisons and hyperboles, but about using language in a way that antagonises and infrahumanises other commentators, as well as party political leaders.

When it comes to political discourse, an interesting result was the fact that while negative sentiment towards the opposition party did not impact the users’ engagement, the positive sentiment towards the government party did in fact negatively influence the probability of users’ reaction. Such a result would suggest that commentators are more polarised in the context of the current ruling party, while the opposition party has less impact on them. When we compare the results of the frequency analysis and the logistic analysis, there can be seen some correlating findings. On the one hand, it was under the government party’s post that more contemporary debate was seen (as researched through word frequencies). At the same time, the quantitative result showed that sentiment towards the ruling party was influential in user engagement. Therefore, it can be implied that the ruling party is more involved in incorporating elements of historical heritage in their social media content and that readers of such content are replying in theme. The frequency analysis had also spotlighted the fact that there is no unique way in replying to social media content that relates to historical events. In such, we add to the existing literature by showcasing different ways readers interact with posted content and comments following it.

The given results have shown that there are certain antecedents of user engagement with comments within the context of political discourse. Firstly, the quicker comments were appearing under political parties’ posts, the more engagement from users they have seen. Our findings support those of M. A. Xenos, Macafee, and Pole (2017) that longer visible content is usually more frequently seen by other users. In this sense it can be argued that users interacting with political content usually limit their readings to the freshest items and when the quantity of comments is increasing, they do not go on to read all of the texts. What is more, the use of “graphicons” also turned out to be a significant factor in generating user engagement with comments, but its effect was of a negative nature. In this context, we have filled in the gap showed by Herring and Dainas (2017) and showed that GIFs or pictures do not augment the communication in political posts’ comments. It can be suspected that some of the posts that consisted only of “graphicons” do not convey enough content for users to deem them interesting enough to interact with. On the other hand, when it comes to the length of the comments, the length of the text did not significantly impact how people reacted to it. Such a result stands in opposition to the findings of Filieri (2015) and shows that patterns of interaction with user generated content are different in a political and leisure context. However, the subject of this difference needs more in-depth investigation in future research. The given findings also support the conclusions of Platonov and Svetlov (2021) that the readability of comments did not influence the reader engagement. In the current case, the similarity of comments’ understandability might be attributed to the fact that comments are usually short in length (which as mentioned also did not influence engagement). When it comes to sentiment, the gathered data showed results in line with those of Alashri et al. (2016) and partially opposing the findings of Segesten et al. (2020). In the case of political posts about the 1981 Martial Law, the comments that were of a positive sentiment towards it received significantly fewer reactions. This would suggest that people tend to ignore comments which go against their beliefs, at least in the context of historical events.

Conclusion

The given research further proved that the Martial Law of 1981 is still a burning issue in contemporary Polish political discourse. More than 70% of people who took part in the parliamentary elections in 2019 voted for parties that posted information about martial law and its victims on their Facebook profiles. The attitude of political parties to Martial Law is ambiguous, because the parties that decided to publish posts on this day were critical of it or worshiped its victims (Law and Justice and Civic Platform), but there were also parties that did not publish anything on the anniversary that could be associated with the introduction of Martial Law in 1981 (the Left and Polish People’s Party). Such differences are likely due to the parties’ genesis. Additionally, in the case of political parties, one can speak of a resignation from forcing one's attitude towards the history of Martial Law. It further shows that political discourse differs in the traditional media and social media.

The attitude of the commentators to the Martial Law introduced in 1981 under the posts of political parties is negative, as is the attitude of the commentators towards political parties, although more so towards the ruling party (Law and Justice). Interestingly, the sentiment in the comments about Martial Law has decreased both under the comments of the Civic Platform (2020 and 2021) and Law and Justice (2020 and 2021), even though the general tendency is the opposite – Martial Law is assessed worse and worse in society over the years (Pankowski 2021). Moreover, the negative attitude towards Martial Law in the comments is greater under the Civic Platform posts than under the Law and Justice posts – it is puzzling as it is Law and Justice voters who are more critical of Martial Law as such (Pankowski 2021). The first explanation for this relationship would be the fact that commentators on Facebook are not a representative group. The second, perhaps compatible with the previous one, is that the posts of a political party are commented on not only by its supporters, but also by those who do not sympathise with it. This is possible and the results of our research would suggest it – a greater percentage of comments on Civic Platform posts are unfavourable to the Civic Platform than Law and Justice posts, and a greater percentage of comments on Law and Justice posts is unfavourable to Law and Justice than Civic Platform posts. In the comments under the posts of political parties, particular attention is drawn to the significant difference in the content of the comments (including keywords). It should also be noted that the comments were written in a very understandable way, and the Martial Law of 1981 itself was compared (and sometimes even considered as milder and better) with the current socio-political situation in Poland: the permissibility of abortion, restrictions (or the lack thereof) related to the Covid-19 virus. At the same time, readers’ interaction with the content was highly connected with the time of its publication, suggesting that interaction with political posts is limited in time and usually happens on the day of publication. Interestingly, the length of the comment was not an antecedent of interaction, which would imply that it was the content quality not the quantity that pushed users to interact with the comment. Such an implication can be also derived from the fact that posts containing graphicons usually generate less interest from the readers.

Limitations and future research

It should be remembered that the comments of Facebook users were analysed only under four posts by two political parties on December 13, 2020 (Law and Justice and Civic Platform) and December 13th, 2021 (Law and Justice and Civic Platform) – this does not therefore give the right to infer as to the general attitude of political parties to Martial Law, let alone the attitude of Poles, because those who comment on posts on social networks are not a representative group for the general public. At the same time, due to the focus on the political parties discourse, certain SU points of view were limited in the analysis. It should be also emphasised that only Facebook was analysed and this is not representative of the entire social media debate. In this sense, our study has also shown concerning aspect of short timeframe of interaction with political media content. However, it does not limit the possibility of using it as a political debate framework as it is a reflection of modern style of interaction with online content and opens room for future research in this subject. Although the collected data do not provide answers to the attitude of Poles to Martial Law, they indicate the opinion of those who comment on the posts of political parties. It would be very interesting to investigate who decides to comment on the posts of political parties on Facebook.

It would also be interesting to conduct a similar study of the posts of political parties and comments under them on the anniversaries of other events in the recent history of Poland that evoke extreme emotions: (1) partially free elections (June 4th 1989) and the fall of Jan Olszewski's government (June 4th, 1992), (2) Poland's accession to the European Union (May 1st, 2004) and (3) the TU-154 plane crash in Smolensk (April 10th 2010). Finally, more research is needed to confirm which elements of user comments attract the attention of other readers, as the given study has showed certain differences in the context of political content.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The research has been supported by a grant from the Priority Research Area Heritage under the Strategic Programme Excellence Initiative at the Jagiellonian University.

Notes on contributors

Jakub Krupa

Jakub Krupa is a PhD student in field of Political and Public Administration Sciences at Jagiellonian University. His research areas of interest are mainly connected with media bias and populist discourse.

Adam Jezierski

Adam Jezierski is a PhD student in field of Management and Quality Studies at Jagiellonian University. His research areas of interest are mainly connected with tourism and hotel management with special focus on reputation management in online environment. He also possess practical experience both at working and managing hotel properties in Europe and Asia.