Critical Discourse Theory
In contrast to more traditional lines of inquiry localized to the political
sciences, this project employs critical discourse analysis methods in its analysis
of speech acts concerning immigration during the 2016 presidential campaign season.
Rather than analyze the potential implications of platforms proposed by presidential
candidates, this projects concerns itself with the linguistic underpinnings that
necessitate such platforms. We argue that vastly different immigration and
immigration reform proposals put forth during the 2016 presidential election cycle
were rooted in as much, if not more, a given candidate's immigration rhetoric as
opossed to any so-called fact based reality. Rather, in keeping with the post-modern
traditions of
Michael
Focault, we aim to demonstrate that knowledge and reality are constructs of
language, rather than objective truths . Thus, advocacy for building a wall along
the United States' southern border can be a viable platform for constituencies that
use the phrase "illegal immigrant"- just as amnesty and a pathway to citizenship can
be fitting for those who use the terms "dreamer" and "undocumented."
Political Discourse and the 2016 Election Cycle
Defining exactly what constitutes political discourse can prove to be its own
endeavor. For simplicity's sake, we find van Djik's definition of political to be
most poinent: "the easiest, and not altogether misguided, answer is that political
discourse is identified by its actors or authors, viz., politicians."
-Teun van Djik Nevertheless, we feel it necessary to explicate further our
decision to analyze immigration discourse from the 2016 presidential election cycle.
Without a doubt, rhetoric from the 22 analyzed presidential candidates running for
election constitutes political discourse. However, we will argue that the 2016
presidential election campaign season marked a moment of inflated political
participation- both in terms of candidates and constituents. On the campaign trail
candidates engaged in a considerable number of discursive acts including speeches,
rallys, town-halls, and debates. At the same time, a record number of recipients
attended, listened, and viewed these speech acts. Thus, we contend that the
heightened degree of political participation during the 2016 presdiential election
cycle, on account of politicians and constitutents, marks an appropriate moment to
analyze political discourse on immigration.
Concerning Method
In our XML markup phase we focused on two levels of rhetorical analysis. On
one level, tagged individual words that were relevant to immigration rhetoric. Words
describing immigrants, illegal/undocumented immigrants, or refugees recieved the tag
"immigrant" with attribute values "status," denoting the subject's legal immigratory
status and "term" denoting the actual term used by the candidate. Immigration
related terms recieved identical treatment, however under the tag "immigration"
instead. In addition we included "keyword" tags, which were used on individual words
or phrases pertaining to immigration such as amnesty, border, wall, pathway to
citizenship, or Dreamer.
On the second linguistic level, we marked what we
called "tropes." Trope tags, as inspired by the "frames of reference" described in
Dr. Luke Peterson's Israel Palestine in the Print Media, were used to
describe any speech utterances that disucssed immigration in terms of a specific
ideological positions, or frames of reference. These ideological dispositions are
often rooted in a candidate's definition of a given issue, and propogate an
emotional rather than objective or fact based response. Thus, utterances such as
"the fact is, since then, many killings, murders, crime, drugs pouring across the
border, are money going out and the drugs coming in," employs a homeland security
frane of reference. By this perspective illegal immigration should be viewed as a
harbinger of crime, drugs, and death to innocent Americans. Other tropes employ
economic and American value based statements in order to define immigration in a
positive or negative light. In our markup, we tagged sentence similar to our example
denoting the trope's "type" grouped as economic, national security, and American
values. Then, when applicable these tropes were given a subtype to further specifiy
an observed fram of reference. Thus, the latte example would be tagged with the
"trope" element, "secure" type, and "crime drugs" subtype denoting a frame of
reference that depicts immigrants as criminals and drug pushers.
Topic Modeling with MALLET
Along with our critical discourse analysis conducted through manual textual
markup, a statistical textual collection and analysis tool called MALLET was
implemented to identify "topics" in the presidential debates. These "topics" are
generated entirely by the program, by feeding it the text of each debate, at which
point the program identifies words that appear in close proximity repeated times
throughout the discourse. Once the word clusters are captured by the program, it
then assembles those word groups into lists which become respective "topics." It is
worth mentioning that, while the topics created by the program can produce coherent
textual themes that manual analysis could not produce, the program itself
(obviously) has no idea what the words it selects actually mean. That is to say, the
words themselves are nothing more than a string of letters to the computer, who
assembles lists on that basis alone. The analysis of its topic collection and the
significance of each of the topics in regard to our central research questions is
where we, the researchers, re-entered the equation.
Collocation Analysis
Similar to topic modeling, collocation analysis is another corpus-level
textual analysis tool that identifies regularities in word usage in a body of texts.
Unlike topic modeling, however, collocation analysis does not generate topics, but
rather it simply collects the words that co-occur in greatest frequency in a
selected body of text. In the context of our research, the collocation elicited
terms that appeared in tandem in the immigration sections of the debates.
Sources
A special thanks goes out to Gerhard Peters, John T. Wooley, and their
American Presidency Project, whose debate transcripts we have borrowed for this
project. We highly recommend users to take a look at their project. They can be
found at
http://www.presidency.ucsb.edu