Protesting With Feeling in Puerto Rico: Twitter and El Verano Del 19

• Published date: 01 June 2023
• DOI: http://doi.org/10.1177/10659129231161286
• Author: Jason Gainous,Mayra Vélez Serrano,Kevin M. Wagner
• Date: 01 June 2023

Article

Political Research Quarterly

2023, Vol. 76(2) 465–480

© The Author(s) 2023

Article reuse guidelines:

sagepub.com/journals-permissions

DOI: 10.1177/10659129231161286

journals.sagepub.com/home/prq

Protesting With Feeling in Puerto Rico:

Twitter and El Verano Del 19

Jason Gainous

1

, Mayra V ´

elez Serrano

2

and Kevin M. Wagner

3

Abstract

Communication research has extensively addressed the inf‌luence of social media on protest. We seek to add to this body

of research by examining how sentiment contained in Twitter communication about protest can condition the public

reach of this communication. Specif‌ically, we are interested in whether Twitter communication cou ched in negative

sentiments like anger, disgust, and fear will drive this reach more than communication couched in positive sentiment such

as trust, joy, and surprise. We rely on sentiment analysis to examine roughly 30,000 tweets surr ounding the summer of

2019 Protests in Puerto Rico. These protests centered on a scandal involving former Governor Ricardo Rossello that

ultimately led to his resignation. The analysis required adapting an English language sentiment dict ionary to Spanish. Our

results suggest that protesters frequently tweeted with both positive and negative sentim ent when calling for the

governor’s resignation, but ultimately, those tweets couched in negative sentiment, when compared to those with

positive sentiment, had the most reach. That said, those tweets including either positive or negative sentiment had more

reach than those absent sentiment.

Keywords

protest, Twitter, sentiment analysis, Puerto Rico

There has been a great deal of research focused on how

social media can shape protests and protest behavior.

However, there is opportunity to better understand how

the actual content, the words used, in social media

communication can guide the effectiveness of protest

communication. In the summer of 2019, Puerto Rico’s

governor was embroiled in a scandal. This resulted from

the public release of a series of private conversations on

Telegram between the governor and his colleagues. In

these conversations, the former governor’s comments

were homophobic, discriminatory and mocked the victims

of Hurricane Mar´

ıa. Mass protests ensued, and ultimately,

Governor Ricardo (Ricky) Rossello resigned. These

protests were both digital and traditional, but even those

traditional in-person protest events were largely organized

via social media. This political event was both relatively

recent and signif‌icant. It resulted in a largely unprece-

dented outcome: the resignation of a Puerto Rican

Governor. As such, we are particularly interested in how

the nature of the digital communication here may have

shaped the spread of dissent, and the call for the former

governor’s resignation.

We theorize that political communication is larger than

just the idea being communicated. It should be seen more

broadly in the context in which these ideas are delivered.

The emotional delivery of these ideas may condition their

eff‌icacy and spread. As psychologists, market researchers,

and life coaches might put it: it’s not just what you say, but

how you say it. In this research, we examine the effec-

tiveness of social media inf‌luence by considering the

sentiment used in the message. In particular, we categorize

and measure the sentiment contained in Twitter com-

munication including the popular hashtag #rickyrenucia

that called for the governor’s resignation. We test whether

tweets couched in negative sentiments like anger, disgust,

and fear drove the spread, or reach, of that communication

more than those couched in positive sentiment such as

trust, joy, and surprise.

1

University of Louisville, Louisville, KY, USA

2

University of Puerto Rico, San Juan, PR, USA

3

Florida Atlantic University, Boca Raton, FL, USA

Corresponding Author:

Jason Gainous, Department of Political Science, University of

Louisville, Ford Hall Room 406, Louisville, KY 40292, USA.

Email: jason.gainous@louisville.edu

While sentiment analysis that uses constructed lexi-

cons or dictionaries of words known to represent different

emotions to count and characterize the emotion in com-

munication is becoming quite common, there is now

growing research that conducts such analysis in languages

other than English Mohammad and Turney, 2013.We

seek to expand that approach here by constructing a

Spanish language dictionary. Instead of reinventing the

wheel, we decided to translate a popular existing dic-

tionary, the NRC Emotion Lexicon. After a descriptive

exploration of the corpus of words used in a sample of

over 30,000 tweets containing #rickyrenucia, we examine

the distribution of sentiment in those tweets. Concluding,

we narrow the population of tweets down to those in-

cluding at least one word from our sentiment dictionary

(n= 11,451). Then, using these emotionally charged

tweets as data, we f‌it a series of negative binomial re-

gressions to estimate the relationship between both dis-

crete emotions and emotional valence in those tweets with

the reach of those tweets. We measure reach using a series

of characteristics about the individual tweets including

how frequently they were favorited, retweeted, and re-

plied to among other things.

Our results are quite clear, those tweets with strong

negative emotional appeals are far more likely to drive

reach than those with strong positive emotional appeal.

Before laying out our empirical strategy in detail, and

walking through the results, we provide the theoretical

frame. Here, we seek to provide a lens to understand how,

and why, emotional appeals may be more effective than

those lacking emotion. Further, we offer a discussion of

the context in Puerto Rico. This discussion makes clear

why emotions ran hot in El Verano del 19 (The Summer of

2019).

Saying It With Some Feeling…

Information and the effect is has on the formation of

attitudes and behaviors has long been modeled by scholars

(for a few see Garrett 2009,Sears & Freedman 1967,

Zaller 1992). Signif‌icant work has been done on how

protests spread with f‌indings showing the importance of

communication, political interest and organizational ties

(Schussman & Soule 2005). Indeed, long before social

media, protest diffusion was predicated on effective or-

ganization and activists (Andrews & Biggs 2006). The

rise of the digital information sphere has altered some of

the variables that may affect the character and nature of

this process. Traditional paradigms of journalism and the

manner and nature of media consumption have shifted

with the expansion and dominance of online digital and

social media networks. Consuming news online changes

two foundational parts of social information structures.

The entrance into the media sphere is far easier, and the

legacy media can no longer act as a gatekeeper restricting

the supplier of political news. Second, the consumer is an

active participant in def‌ining the market for news. Indeed,

rather than be a passive receiver, the consumer has the

ability to def‌ine, limit, and even respond to the media

platforms (Gainous & Wagner 2014).

Much of this is the product of patterns of online news

consumption. When given the opportunity to choose what

news to read or view, people can use online platforms to

tailor their news feed so as to avoid the discomfort caused

by having their beliefs challenged by ideas, facts, or ar-

guments that contest their settled assumption and long-

held opinions (Gainous & Wagner 2014). Alternatively,

other scholars have found that rather than avoiding dis-

agreeable information, many people are just selectively

seeking out content that supports their established views

(Garrett, Carnahan & Lynch 2013). Regardless, the end

result is the formation of information structures that are

receptive to certain types of messages and messaging.

Further, while there is no gatekeeper, the position of the

users on the social network matters. A small core of

central users can generate signif‌icant message diffusion

based in their location in the network (Gonz´

alez-Bailón

et al. 2011).

This presents an opportunity for off‌ice-seekers and

other types of political actors. They can target messages to

groups formed around particular beliefs and ideas while

avoiding the gatekeeping function of the legacy media.

Politicians need and want the opportunity structure, and

the consumers of news are seeking a preferred stream of

information that political actors can provide (Garrett,

Carnahan & Lynch 2013). The end result is that digital

media allows both the political actor and the news con-

sumer to create mutually benef‌icial media spheres where

the object of the news contributes to their own coverage.

Further, the news platforms online may even inten-

tionally tie themselves to the political actor to gain the

page views provided by the politician themselves (Wagner

& Gainous 2021). The economics of modern news

encourages legacy media to engage with candidates,

off‌ice-holders, and online political inf‌luencers. Online

publications will ref‌lect and reinforce the politician’s

messages as it gains page views and drives readership

(Wagner and Gainous 2021). Online media can drive large

numbers of page views that increase media revenue

(Karpf 2016). The entire system is reinforcing. The tra-

ditional media reports on social media, and the digital

platforms disseminate the stories.

Controlling the message is not a new political strategy.

It is just accomplished far more easily with advancing

digital technology. Indeed, the shaping of public opinion

by political actors is well-studied in the areas of agenda-

setting and framing (Graber 2009). Elite public opinion

formation holds, very basically, that elites—elected and

466 Political Research Quarterly 76(2)