Stance Taking in Social Media: the Analysis of the Comments About Us Presidential Candidates on Facebook and Twitter

Roma Kriaučiūnienė

Vilnius University, Faculty of Philology, Institute of Foreign Languages, Lithuania
roma.kriauciuniene@flf.vu.lt
ORCID iD: 0000-0002-9356-1098 https://orcid.org/0000-0002-9356-1098

Jefferey La Roux

Vilnius University, Faculty of Philology, Institute of Foreign Languages, Lithuania
jefferey.la@flf.vu.lt

Miglė Lauciūtė

Vilnius University, Faculty of Philology, Lithuania
migle.lauciute@gmail.com

Summary. The subject of the paper is the analysis of the expression of stance taking in an online environment, mainly in the comments of users of social networks such as Facebook and Twitter about the presidential candidates of the American Presidential Election in 2016. The empirical data analysis was carried out following the ideas of J. W. Du Bois (2007), D. Barton & C. Lee (2013) and R. Englebretson (2007) on stance taking and J. W. Du Bois’ (2007) model of stance triangle, i.e. grouping instances of stance-taking into one of these groups: evaluation, affect or epistemicity, which served as the main framework of this study. The work of linguists D. Barton & C. Lee (2013) on the expression of stance-taking in an online environment were also taken into consideration.

Having in mind the fact that stance identification is a challenging task , i.e. it could be implicitly as well as explicitly expressed and that it should be inferred from different modes of its expression and interpreted with reference to many contextual and intertextual factors, in the current analysis the authors focused on interpretation of linguistic as well as other multimodal means of the expression of stance that were used by users of social networks in their writing spaces on the topic of the Presidential Election in the United States in 2016. It should also be mentioned that the analysis presented in this article offers only one of the many possible interpretations of the data. Moreover, the current paper concentrates mainly on the presentation of the empirical data of the expression of affective stance. However, it should be indicated that in some cases stance types overlap, i.e. one instance could be treated as both taking an affective and an evaluative stance, as judgements and evaluation (i.e. evaluative stance) are often based on feelings (i.e. affective stance).

The main source of the empirical data were the instances of stance taking taken from comments found on Donald Trump’s and Hillary Clinton’s verified Facebook and Twitter pages during their presidential campaigns in 2016. All in all, 147 examples of posts and comments from the social networks Facebook and Twitter were collected: 72 comments incorporating stance taking on Donald Trump‘s posts, and 75 comments including stance taking on Hillary Clinton‘s posts. The results of the empirical data analysis showed that the affective stance was expressed by linguistic as well as multimodal means.

Key words: stance taking, affective stance, social networks, comments, online interaction, multimodal means.

Copyright © 2019 Roma Kriaučiūnienė, Jefferey La Roux, Miglė Lauciūtė. Published by Vilnius University Press
This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Article submitted 17 January 2018, accepted

Roma Kriaučiūnienė is professor of the Institute of Foreign Languages of the Faculty of Philology of Vilnius University. Her professional interests include media linguistics and educational sciences (expression of stance, multilingualism and plurilingualism, foreign language didactics, value attitudes). Her research papers were published in Lithuania, Latvia, Russia, Turkey, Slovenia, Italy, India, Check Republic, Spain.

Jefferey La Roux is an English language lecturer at Vilnius University. He has worked as an English teacher in the United States. He also has a degree in American history and enjoys following American politics and trends. His research interests include dialectal differences, words perceived to be banned, and the changing meanings of words.

Miglė Lauciūtė is a graduate of the first cycle study programme “English and the other Foreign Language (Spanish, French, Norwegian)“ of Vilnius University, Faculty of Philology.

Introduction

The pre-election period in the USA attracted the attention of many people around the world and especially of those readers and viewers of news who are active on social networks such as Facebook and Twitter. Usually the comments on social networks are far from being neutral and objective. On the contrary, people are very open in their comments and reveal their stance towards any subject under discussion very freely. As D. Barton and C. Lee (2013, p.86) state: “Writing spaces in new digital media not only offer opportunities for multilingual text representation, they also serve as new domains for people to express their opinions and attitudes on many topics, alongside traditional modes of communication such as face-to-face talk and written texts.“ Thus, the way comments are written by using different linguistic and multimodal means of communication(no comma here) can reveal the opinions of electorates, i.e. their personal feelings, value attitudes, beliefs, assessments and judgments and can shed light on the values of entire communities. In other words, people express their stance towards what they are writing about. Moreover, stance is interactional in nature and as Englebretson (2007b) states, is collaboratively coming into being among the participants in an exchange and/or by virtue of opposition to other stances. Thus, comments on social networks become a subject for evaluation by others, can influence other people’s opinions and consequently have an impact on their judgement.

According to D. Barton and C. Lee (2013, p.87), (comma should be placed after the parenthesis, not after stance, and let’s italicize stance so the reader knows we are talking about the word itself.) stance refers to the position people take in relation to oneself, to what is said, and to other people and objects. The authors state that in any given stance statement, there are three major components – the person expressing the stance, the topic being discussed and the resources being drawn upon. Besides, stance statements are always directed towards a particular audience, therefore the addressee might be considered to be the fourth element of the stance taking process.

There are numerous ways of taking a stance, especially in an online environment – from the choice of words to clicking a particular button on a website and by employing different modes of expression. For example, when writing a post in social media in addition to a written text, i.e. a linguistic mode, one uses some audio or visual representations as well. Selecting one mode over another can also be interpreted as taking a particular stance. As D. Barton and C. Lee (2013, p.88) put it, stance taking online is no longer a solely monomodal linguistic act: “In addition to a written word, many WEB 2.0 media provide built-in features that invite people to express opinions publicly through multimodal means.“ There is a possibility of rating the posts online, and in that way take a stance. For instance, YouTube provides thumbs up and thumbs down buttons and a comment section that allows viewers to express their stance on a particular aspect of a video, Facebook has the famous “Like” button, Flickr members can mark certain photographs as favourites, etc. A famous phrase coined by Marshall McLuhan basically states the same: “The medium is the message” (Marshall McLuhan, 1967). The way one chooses to convey a message can carry a certain meaning or message within itself. Or in other words, the manner in which information is communicated has more power and is more important than the information itself. Thus, stance taking is a prevailing phenomenon in any kind of communication and there is a wide array of linguistic features and other modes of expressing it.

It should be mentioned that the type of discourse, where stance taking can be observed most, is political discourse, which contains one or another kind of stance taking. It is particularly evident, for example, when a politician expresses an idea for new legislation and in a way takes a stance. Moreover, a great deal of political campaigning, advertising, debating and discussing opinions are expressed on various social media platforms. In this article, political discourse in social media is treated as the comments and posts on the social media platforms Facebook and Twitter. The main topic of the comments and posts is the 2016 Presidential Election in the United States.. The posts were written by two candidates, Hillary Clinton and Donald Trump, and the investigated comments, which is the subject and the main focus of this paper, were composed by ordinary Facebook and Twitter users who took an interest in this precise topic.

Stance has become an important concept within linguistics and there has been a considerable amount of research carried out in the area of stance. Many scholars have been interested in specific perspectives of stance. For example, V. Simaki, P. Simakis, C. Paradis, and A. Kerren (2018) investigated stance-related characteristics in a data set from Twitter and Facebook and extracted various corpus-, quantitative- and computational-based features that proved to be significant for six stance categories (contrariety, hypotheticality, necessity, prediction, source of knowledge, and uncertainty). Their research proved that stance identification is a challenging task. In P. Sobhani’s (2017) research the task was carried out to determine stance from the text automatically, i.e. whether the author of the text is in favor of, against, or neutral towards a stance object. Another researcher, K. Hyland (2005, p. 174), specifically focused on stance taking in academic writing. Not only did the author analyse the issue of stance but also presented a model of interaction in academic discourse.

The research of different scholars into stance taking shows that there are many approaches through which stance taking could be analysed, but using stance to understand language in new media is a very recent direction. The current research makes a modest attempt to contribute to the analysis of stance in online interactions. Having in mind the fact that stance identification is a challenging task as it could be implicitly as well as explicitly expressed, it should be inferred from different modes of its expression and interpreted with reference to many contextual and intertextual factors. In our analysis we focused on interpretation of linguistic as well as other multimodal means of the expression of stance that were used by users of social networks in their writing spaces on the topic of the 2016 Presidential Election in the United States. It should also be mentioned that our analysis presented in this article offers only one of the many possible interpretations of the data.

Thus, the aim of the article is to analyse stance taking in social media discourse, i.e. comments on Facebook and Twitter referring to the posts written by two candidates, Hillary Clinton and Donald Trump, on the topic of the 2016 US Presidential Election.. The aim was specified by the following research objectives:

1. To determine the means of expression of affective stance in the comments written by users of Facebook and Twitter about Hillary Clinton’s and Donald Trump’s posts in social media platforms on the topic of the American Presidential Election in 2016;
2. To identify the multi-modal features dominating in the social media comments referring to the posts by Hillary Clinton and Donald Trump written by social media users on the topic of the American Presidential Election in 2016.

The main source of data for this research were social media websites such as Facebook and Twitter.

1. Theoretical background of the research. The notion of stance

Generally speaking, stance is an attitude, an opinion or a belief on a certain topic, especially when expressed publically. C. Hernández-Guerra (2016, p. 132) proposes quite a simple and straight-forward definition of stance, stating that “The term ‘stance’ refers to the expressions, registers and elements that reflect the attitude of the speaker towards the message he/she is transmitting.” Another researcher, M. Al-Shunnag (2014, p. 71), states that people inevitably evaluate entities and propositions they encounter on some kind of level. When talking, a speaker always provides his or her personal stance on a topic. As the author suggests, stance is usually understood in a particular context, otherwise it may lose its meaning. Moreover, the author states that subjectivity-free discourse does not exist, therefore, stance is included in every text at least on a certain level. D. Biber and E. Finegan (1989, p. 92) state that stance is “the lexical and grammatical expression of attitudes, feelings, judgments, or commitments concerning the propositional content of a message.” Considering the definition, Englebretson (2007, p. 16) proposes that stance can be subdivided into three separate groups:

1. Evaluation (“judgments”)(Notice here we have the spelling of judgments)
2. Affect (“feelings”)
3. Epistemicity(“commitment”)

J. W. Du Bois (2007) assures us that one of the most essential things people do with words is express their point of view. Stance is involved in every type of discourse; it is inevitable to produce some kind of interactive text or speech without taking a stance. Du Bois (2007, p. 139) proposes that a stance can be understood as a “linguistically articulated form of social action”. The other thought expressed by the same author is that stance cannot be profoundly and fully interpreted if one does not consider it in a larger dialogical context. Du Bois (2017, p. 162) also proposes the idea of the Stance Triangle which has three main components: evaluation, positioning, and alignment. Evaluation is the most commonly discussed component, which means the characterization of an object of stance as having a particular quality or value for a stance taker. Positioning can be understood as taking an affective or epistemic stance. Affective stance means that a person chooses to express his stance with an affective predicate, such as “I am glad” provided in the example below. Epistemic stance, on the other hand, is related to one’s knowledge rather than feelings. It either indicates one’s knowledge or ignorance. Alignment can usually be interpreted as the agreement between two stances or two stance takers. It can be expressed with words like agree but stance markers like yes or no, and gestures like a nod or a handshake are more common. The other important factor as indicated by Du Bois (2017, p.162) is that most stance-taking utterances are formed using the first person pronoun I as the subject. Moreover, instances of examples of each stance type are provided by the author Du Bois (2017, p. 162) in the following way:

Evaluation: “That's horrible.”

Positioning: “I'm glad.” (Affective stance) or “I don't know.” (Epistemic stance)

Alignment: “I totally agree with you.”

However, not all scholars agree on such subdivision of stance and propose yet another one. M. Charles (2007) for example, who elaborately investigated the notion of stance from a semantic point of view, distinguishes epistemic stance and attitudinal stance and provides this definition: “epistemic stance marks the status of information in terms of its level of certainty or factuality, and attitudinal stance that indicates the writer´s personal feelings or opinions about a proposition” (2007, p. 206).

Hyland (2005, p. 175) expressed an interesting idea concerning evaluation in that people tend to evaluate things, or take a stance, according to a particular standard; it can either be a communal belief, ideology or anything else that people can relate to. The author goes on to mention other concepts that are crucial for evaluation. He stated that there are two ways to produce an effective argument. The first of them is stance, which is “a textual voice” or community recognized personality (Hyland, 2005, p. 176). To put it in other words, it is the author’s opinions, judgments and commitments or the opposite of that – the author’s choice to conceal his involvement. The other important phenomenon, which helps to evaluate, is engagement, which means the author’s ability to connect to the audience, recognize them as a part of discourse and help them with interpretation. It should also be added that I. Vandergriff (2012, p. 54), for instance, explores yet another interesting concept of metastance, i.e. taking a stance on a previous stance, which is particularly evident in online interaction. Furthermore, the author considers not taking a stance as a kind of stance, thus there is no way to avoid taking a stance.

In addition, it is necessary to research how stance taking occurs in an online environment. According to D. Barton and C. Lee (2013, p. 88), the most significant difference is that stance taking online can be multimodal. In opposition to what has been discussed before, these authors explain how stance can be taken not only through language but also through images. It is impossible to be a part of the online community without taking a stance as almost everything done online from clicking the ‘like’ button on Facebook to leaving a comment on a certain article, indicates a particular point of view, even though it is not always very clear. These authors provide a more simplified definition of a stance. “Stance refers to the positions people take in relation to oneself, to what is said, and to other people or objects.” (Barton, Lee, 2013, p. 19). There are three main components that are involved in stance taking process: the person stating a stance, the topic which is discussed and the resources used. They also add the fourth factor, which is the addressee, that can influence the stance taker. Another idea expressed by D. Barton and C. Lee (2013, p. 87) is that selecting a certain language to take a stance can itself be a stance. It especially applies to bilingual or multilingual people because they have the possibility to express themselves in various languages.

2. The expression of stance and stance markers

B. Gray and D. Biber (2014, p. 220) indicate that stance can be transmitted through evaluative or value-laden lexis, which in most cases is almost implicit. To illustrate this idea the authors provide adjective examples of happy, angry, and surprised, which indicate attitudes or emotions (i.e. affective stance) and other adjective examples including good, wonderful, and lovely, which mark one’s evaluation (i.e. evaluative stance). According to B. Gray and D. Biber (2014, p.221) the other way to express stance is by using adverbials. Stance, expressed by adverbials, is usually very clear and straight-forward, for instance, obviously, in fact, surprisingly. In addition, modals and semi-modals (e.g. may, has to) are also considered to be stance markers. However, they are considered to be less overt because they form a part of a verb phrase. The next method to express stance is by using stance nouns that are also thought to be quite obscure and difficult to notice. These nouns are usually followed by a prepositional phrase (e.g. importance of, hope for, etc.). However, it was noticed that the prepositional phrases frequently do not clearly express “proposition”, as the authors name it. As a result, stance nouns are often ambiguous and difficult to detect. Their other feature is that they are rather ambiguous in nature, because they do not make an explicit reference to the writer/speaker, therefore it could be quite unclear if the stance belongs to the speaker/writer or if it is a general belief.

B. Gray and D. Biber (2014, p.228) provide the following stance noun and adjective framework:

1. Stance adjectives + that-complement clauses;

2. Stance adjectives + to-complement clauses;

3. Stance nouns + that-complement clauses;

4. Stance nouns + to-complement clauses.

3. Methodology of the research

The empirical data analysis was carried following the ideas of J. W. Du Bois (2007), D. Barton & C. Lee (2013) and R. Englebretson (2007) on stance and J. W. Du Bois’ (2007) model of stance triangle, i.e. grouping instances of stance-taking into one of these groups: evaluation, affect or epistemicity. The work of linguists D. Barton & C. Lee (2013) is related to stance-taking in online environment, and its main differences, as they state, is that stance online can be multimodal.

As the article aims to research stance taking in political discourse, particularly the language in social networks during the Presidential Election in 2016, the ideas of T.A. Van Dijk (1997), one of the main researchers of political discourse, were also taken into consideration while conducting an empirical research. As T. A. Van Dijk (1997, p. 11) suggests, the political discourse (text and talk) is produced by politicians and other political institutions (such as presidents, prime ministers, etc.). Thus, important components of political discourse are the authors, who produce texts and talk and the receiving end, i.e. the audience, the people, the citizens, etc. Furthermore, according to T. A. Van Dijk (1997, p. 14) there is the context, including communicative events and encounters, which plays a huge role in political discourse. Additionally, such factors as time, place, circumstance, occasions, intentions, functions, goals, and legal or political implications should also be taken into consideration. This paper concerns the previously mentioned topic of the presidential election in the US, therefore, the context in this research is of great importance.

The empirical analysis of the expression of stance taking in social media was carried out in the following way: firstly, the examples of comments (written by the users) on the American Presidential Election in 2016 about social media posts (written by two candidates Hillary Clinton and Donald Trump) were gathered. Then the type of stance was determined and multi-modal features prevailing in the comments (written by the users) found in the collected examples were identified. However, it should be indicated that in some cases stance types overlapped, i.e. one instance could be treated as both taking an affective and an evaluative stance, as judgements and evaluation (i.e. evaluative stance) are often based on feelings (i.e. affective stance). However, the current paper will focus only on the presentation of the empirical data of affective stance.

Qualitative research method as well as descriptive and analytic approaches were applied when conducting a detailed analysis of the collected comments written by users of Facebook and Twitter in reference to the posts written by two candidates, Hillary Clinton and Donald Trump, in order to determine the type of stance of a particular instance. The main source of the empirical data was the social media platforms such as Facebook and Twitter. The instances of stance taking were taken from comments found on Donald Trump’s and Hillary Clinton’s verified Facebook and Twitter pages during their presidential campaigns in 2016. It is noteworthy that the identity of the Facebook/Twitter users whose comments are researched in this paper will not be revealed and has no significance to the findings of the research. Additionally, all of the data taken from those websites is public and easily accessible. Moreover, the difference between a post and a comment should be clarified. A post should be understood as a social media entry written by a candidate (Donald Trump or Hillary Clinton) and a comment is an entry produced by a social media user, proposing a reply to a certain post made by the presidential candidates. Furthermore, it should be noted that all of the comments are original and unedited and therefore may include some grammar or spelling mistakes. All in all, 147 examples of posts and comments on the social networks Facebook and Twitter were collected: 72 comments incorporating stance taking on Donald Trump‘s posts and 75 comments including stance taking on Hillary Clinton’s posts. The analysis of stance taking was carried out according to the division of stance proposed by Robert Englebretson (2007) who determined three stance types: affection, evaluation, and epistemicity. The current paper will focus mainly on the presentation of the empirical data of affective stance and identification of linguistic and multimodal means of expressing it.

4. The analysis of the affective stance in the comments on Donald Trump’s posts on social media platforms

While gathering examples of the comments on Donald Trump’s posts on social media platforms such as Facebook and Twitter, it was noted that 24 examples in total contained instances of affective stance. Some examples of expressing affective stance contained both linguistic and multimodal means (affective verbs, adjectives, emoticons, etc.). In the collected examples, six main means of expressing stance were employed, which are presented below.

1. Expressing affective stance using verbs of emotion such as love, hate, like, etc., which communicate very strong feelings and consequently demonstrate a person’s point of view on a certain topic. Most of the verbs found in the examples were positive: four instances of the word love and two instances of the verb like. Moreover, in several instances these verbs were capitalised, thus strengthening the semantic meaning of the verbs and demonstrating stronger emotion, and the affective stance of the writers’ comments as well.

What is more, some verbs were modified by adverbs making up adverb + verb combinations. Adverbs usually add extra meaning to the phrase, thus help strengthen a stance (e.g. truly fear).

Affective stance expressed by the verbs love, like, stand, hail, fear (either in a negative or positive form) was found in the provided examples below, which were retrieved from the Facebook and Twitter users’ comments.

2) The other means of expressing affective stance in the collected examples from the comments by Facebook and Twitter users on Donald Trump’s posts were adjectives. The following adjectives were found: glad, disheartened, excited, and proud. The adjective glad was among the most frequently used adjectives (three times) found in the comments. Some of these adjectives are even given additional power by adding an adverb, such as so, very or a demonstrative pronoun this. It should also be pointed out that in several cases more than one exclamation mark was used, which could be assumed to strengthen emotional attitude expressed by the writers’ comments too.

3) Expressing affective stance using emojis. In some cases, people tend to only put a small pictogram to represent their feelings. It is common knowledge that a heart emoji represents positive feelings, such as love or affection, or that a thumbs-up signifies approval or agreement. One instance of stance taking using emojis was found in the collected examples:

4) Expressing affective stance using emojis. In some cases, people tend to only put a small pictogram to represent their feelings. It is common knowledge that a heart emoji represents positive feelings, such as love or affection, or that a thumbs-up signifies approval or agreement. One instance of stance taking using emojis was found in the collected examples:

According to the Cambridge Advanced Learner’s Dictionary (2013), the first idiom means feeling very upset, worried, or angry. The second idiom is used to express dislike about something or someone, or one’s opinion about something unpleasant. The expression of the affective stance of the writers’ comments is strengthened by the use of capital letters as well as numerous exclamation marks.

5) Expressing affective stance with the help of adverbs. It should be mentioned here that some of the examples mentioned above also included adverbs that modified verbs and therefore showed the attitude of the writers’ comments on a certain topic. One more instance of adverb usage found in the comments could be added here as well:

The adverb sadly shows one’s disappointment and negative emotions regarding the topic of discourse.

6) Expressing affective stance with letter capitalisation. As it has already been mentioned earlier, using capital letters in comments (alongside the usage of numerous exclamation marks) could be interpreted as an expression of affective stance because it provides considerable emphasis on a selected word or a phrase. It is usually driven by strong emotions such as love, hatred, anger, etc. Most internet users tend to understand capitalisation as shouting. Letter capitalisation was found in 11 comments written by Facebook and Twitter users on Donald Trump’s posts:

Considering the content of the Facebook and Twitter users’ comments on Donald Trump’s posts that include affective stance, most of them express positive stance towards Donald Trump and his ideas, people tend to agree with him and even praise him. The other option is that the authors of the comments demonstrate the contrasting stance towards his opponent Hillary Clinton and show dislike or even hatred toward her.

5.The analysis of affective stance in the comments on Hillary Clinton’s posts on social media platforms

To begin with, it is significant to mention that out of the total 75 examples collected 17 contained the expression of affective stance of the writers’ comments. Having researched all of the collected comments on Hillary Clinton’s posts it could be stated that affective stance is most frequently expressed by four different linguistic means and as well as multimodal means.

1) Expressing affective stance using verbs that transmitted strong emotions, either negative or positive. Admittedly, most of the comments contained positive affective verbs. The examples included these verbs: love, adore, dislike, despise. The verb love was the most common; it was repeated throughout all of the examples six times. There are nine comments which contained affective stance expressed with the help of an affective verb love:

Evidently, the above examples demonstrate that the expression of affective stance by verbs is strengthened by adverbs as well, for example: really dislike, adore you so MUCH .In the latter case the expression of emotion is strengthened by using capital letters. Besides, it should be indicated that capital letters as well as the usage of more than one exclamation mark could also be considered as a means of strengthening the expression of affective stance of the writers’ comments.

2) The other means that were used by the commenters to express affective stance were adjectives. These were usually adjectives that provided information about one’s personal feelings and emotions. In most cases, they communicated powerful messages. The adjectives found in the comments were: proud, glad, impressed and thankful. The comments presented below are the six cases containing affective stance expressed with the help of an adjective:

It should also be mentioned that capital letters were also used alongside the emojis and more than one exclamation mark to strengthen the attitude of the writers’ comments and the expression of affective stance.

3) In addition, in one comment affective stance was expressed using emojis. In this particular instance, multiple small heart images of different colours, which could be interpreted as signs of positive emotions, love, loyalty, etc., were used. The example is presented below:

4. The last means to express affective stance is using capital letters, which has been indicated earlier as well. Capital letters signal the focus on a particular word, phrase or a concept in general. The following are examples in which capitalisation could be considered as a means of expressing affective stance:

As regards the content of the comments, the majority of them (14 out of 17) expressed a positive affective stance towards Hillary Clinton, her actions or policies. Most of the comments were concerned with the debates, and usually, but not always, supported Hillary Clinton’s side. People demonstrated affirmative judgement, support and positive emotions. It could be justified by the usage of such positive verbs as love and adore or adjectives glad or proud, etc.

As the comments were about the posts written by D. Trump and H. Clinton, therefore it was worthwhile looking into the content and the prevalent topics of their posts. They all dealt with the election to some extent, however, there were some notable differences. Taking into consideration the content of Donald Trump’s posts, it is evident that the main matter in his posts is criticism of various situations and people, in particular his opponent Hillary Clinton, ex-president Barack Obama and others. Overall, his posts were quite negative in nature. Hillary Clinton’s posts, on the other hand, contained more positive content. Although she drew social network users’ attention to the reasons it was necessary to condemn Donald Trump, her main focus remained on the issues she advocated in her presidential campaign, such as defending women’s rights, improving the economy, fighting for equality, etc.

An interesting piece of research done by Crockett (2016) reveals similar results. After having analysed over 4000 tweets by Hillary Clinton and Donald Trump, the author makes the conclusion that Hillary Clinton is more positive and polite on Twitter. While 38% of her content was positive and 39% was neutral, Donald Trump’s tweets exposed quite different results. Most of Trump’s negative tweets, 45% of his total production, were insults and criticism. In addition, the 30 most frequently used adjectives were investigated. The results revealed that 60% of Donald Trump’s top-used adjectives were negative, often even offensive, while negative adjectives used by Hillary Clinton made up only 20%.

The other article by Manning (2016) analysed the posts on Facebook and their engagement with the public. The author stated that both candidates had “a healthy thirst for likes and shares” (Manning, 2016). The research mainly focused on the engagement the Facebook posts received. Hillary Clinton acquired 3000 shares per post, while Donald Trump obtained twice as that.

Conclusions

It could be concluded that the content of the comments was approximately equally negative and positive in relation to both candidates. The comments on Donald Trump’s posts were favourable to him; respectively Hillary Clinton had more positive comments on her own posts. However, referring to the posts of the candidates that the users of Facebook and Twitter were commenting upon it might be stated that Donald Trump used a more negative lexicon, sometimes even offensive words or phrases. Hillary Clinton, on the other hand, made an attempt to be more neutral or positive in the majority of her posts.

Having analysed the comments about the posts by presidential candidates on social media in terms of expressing affective stance, it could be stated that the most commonly affective stance was expressed by verbs, which usually transmitted a certain feeling or an emotion, for example, hate, love, adore, etc. The other means that were used by the commenters to express affective stance were adjectives, such as: disheartened, excited. It should be mentioned that adverbs were also found to strengthen the meaning of verbs and adjectives expressing stance. There were also a few instances of the usage of idioms to express the affective stance.

The examples of the comments written by the users about the posts written by Hillary Clinton and Donald Trump were analysed in relation to multimodality. It could be noted that alongside the above mentioned linguistic means to express affective stance there were multi-modal means exploited by the comments’ writers as well. Among the most common ones were emojis, capitalisation and numerous exclamation marks that strengthened the expression of the affective stance of the writers’ comments.