A Corpus-Based Study of the Discourse of Recreational Marijuana Legalization

Abstract

Recreational Marijuana Legalization (RML) is strongly supported in business, legal, and political
circles, but opposed mostly by the medical professions. A nationwide rush to legalize recreational
marijuana (a.k.a cannabis) at all levels is underway in the United States. Proponents claim that
there are millions of dollars to be made in tax revenues that will create jobs and address social
justice concerns. Opponents from law enforcement, medical and related fields highlight the risks
to public health of smoking or ingesting cannabis with high concentrations of delta-9-
tetrahydrocannabinol (THC). This study examines the discourse of RML in texts from two
corpora: a corpus of articles from scientific fields, and a reference corpus consisting of texts from
newspapers, online cannabis news, economic and legal journals. Using the Words tool in
LancsBox 6, word frequencies and keywords were obtained. Next, hundreds of concordance lines
of selected keywords as nodes from texts in each corpus were analyzed and evaluated to see
whether word clusters around or near nodes could reveal hidden facts about the RML discourse.
The results show that slightly more positive discourse patterns characterize texts in the reference
corpus, highlighting revenues to be generated by taxing cannabis industry participants, but
information about regulations and hefty license fees is not publicly shared with voters. Keywords
as nodes in medical research texts yield mostly negative associations, stressing harmful effects on
cognition, motor skills, and adolescents’ brains. This has implications for the dissemination of
important factual and scientific information to the average voter through print and online sources