The corpus is composed of 3000 messages (66915 words) posted on Facebook, Twitter and YouTube2, information networks which see a high degree of interaction between their users expressing opinions and arguments related to matters of current interest (Mancera Rueda & Pano Alamán, 2013).

The corpus is composed of messages related to the management of the recent COVID-19 crisis in Spain, extracted automatically using the ExportComments tool. It contains three sub corpus each composed of: 1000 responses to two tweets posted on March 10th and April 24th on the profile @MinisterioSanidad; 1000 comments on two posts dated May 19th and June 1st on the Ministry’s official Facebook page; and 1000 comments on two videos posted on the YouTube channels of El Mundo (01-03-2020) and El País (14-03-2020), about two press conferences given by the Spanish government to detail the initial handling of the pandemic3. Therefore, it includes the users’ responses and comments on these three networks to similar content originating from official sources. Details of the authors have been omitted from the study for reasons of privacy and the messages have been reproduced as they were posted, without leaving out examples of the written subcode inherent to electronic texts.

4. Results

The functions Wordlist and Ngram of the Sketch Engine program have allowed for the compilation of a list of frequencies for the markers present in the corpus. Table 1 shows those which have a minimum of ten occurrences in the general corpus, organised by social network.

Table 1. Most frequent markers (Total ≥10)