Discourse Analysis
Discourse analysis is one of the most popular qualitative analysis techniques we
encounter at Grad Coach. If you’ve landed on this post, you’re probably interested in
discourse analysis, but you’re not sure whether it’s the right fit for your project, or you don’t
know where to start. If so, you’ve come to the right place.
In its simplest form, discourse is verbal or written communication between people
that goes beyond a single sentence. Importantly, discourse is more than just language. The
term “language” can include all forms of linguistic and symbolic units (even things such as
road signs), and language studies can focus on the individual meanings of words. Discourse
goes beyond this and looks at the overall meanings conveyed by language in
context. “Context” here refers to the social, cultural, political, and historical background of
the discourse, and it is important to take this into account to understand underlying
meanings expressed through language.
A popular way of viewing discourse is as language used in specific social contexts, and
as such language serves as a means of prompting some form of social change or meeting
some form of goal.
Discourse analysis uses the language presented in a corpus or body of data to draw
meaning. This body of data could include a set of interviews or focus group discussion
transcripts. While some forms of discourse analysis center in on the specifics of language
(such as sounds or grammar), other forms focus on how this language is used to achieve its aims.
We’ll dig deeper into these two above-mentioned approaches later.
As Wodak and Krzyzanowski (2008) put it: “discourse analysis provides a general
framework to problem-oriented social research”. Basically, discourse analysis is used to
conduct research on the use of language in context in a wide variety of social problems (i.e.,
issues in society that affect individuals negatively).
For example, discourse analysis could be used to assess how language is used to
express differing viewpoints on financial inequality and would look at how the topic should
or shouldn’t be addressed or resolved, and whether this so-called inequality is perceived as
such by participants.
What makes discourse analysis unique is that it posits that social reality is socially
constructed, or that our experience of the world is understood from a subjective standpoint.
Discourse analysis goes beyond the literal meaning of words and languages
For example, people in countries that make use of a lot of censorship will likely have
their knowledge, and thus views, limited by this, and will thus have a different subjective
reality to those within countries with more lax laws on censorship
When should you use discourse analysis?
There are many ways to analyze qualitative data (such as content analysis, narrative
analysis, and thematic analysis), so why should you choose discourse analysis? Well, as with
all analysis methods, the nature of your research aims, objectives and research questions (i.e.
the purpose of your research) will heavily influence the right choice of analysis method.
Discourse analysis definition
The purpose of discourse analysis is to investigate the functions of language (i.e., what
language is used for) and how meaning is constructed in different contexts, which, to recap,
include the social, cultural, political, and historical backgrounds of the discourse.
For example, if you were to study a politician’s speeches, you would need to situate
these speeches in their context, which would involve looking at the politician’s background
and views, the reasons for presenting the speech, the history or context of the audience, and
the country’s social and political history (just to name a few – there are always multiple
contextual factors).
Discourse analysis can also tell you a lot about power and power imbalances,
including how this is developed and maintained, how this plays out in real life (for example,
inequalities because of this power), and how language can be used to maintain it. For
example, you could look at the way that someone with more power (for example, a CEO)
speaks to someone with less power (for example, a lower-level employee).
Therefore, you may consider discourse analysis if you are researching:
• Some form of power or inequality (for example, how affluent individuals interact with
those who are less wealthy
• How people communicate in a specific context (such as in a social situation with
colleagues versus a board meeting)
• Ideology and how ideas (such as values and beliefs) are shared using language (like
in political speeches)
• How communication is used to achieve social goals (such as maintaining a friendship
or navigating conflict)
As you can see, discourse analysis can be a powerful tool for assessing social issues,
as well as power and power imbalances. So, if your research aims and objectives are oriented
around these types of issues, discourse analysis could be a good fit for you.
Discourse Analysis: The main approaches
There are two main approaches to discourse analysis. These are the language-in-
use (also referred to as socially situated text and talk) approaches and the socio-
political approaches (most commonly Critical Discourse Analysis). Let’s take a look at each of
these.
Approach #1: Language-in-use
Language-in-use approaches focus on the finer details of language used within
discourse, such as sentence structures (grammar) and phonology (sounds). This approach is
very descriptive and is seldom seen outside of studies focusing on literature and/or
linguistics.
Because of its formalist roots, language-in-use pays attention to different rules of
communication, such as grammaticality (i.e., when something “sounds okay” to a native
speaker of a language). Analyzing discourse through a language-in-use framework involves
identifying key technicalities of language used in discourse and investigating how the
features are used within a particular social context.
Discourse analysis definition
For example, English makes use of affixes (for example, “un” in “unbelievable”) and
suffixes (“able” in “unbelievable”) but doesn’t typically make use of infixes (units that can be
placed within other words to alter their meaning). However, an English speaker may say
something along the lines of, “that’s un-flipping-believable”. From a language-in-use
perspective, the infix “flipping” could be investigated by assessing how rare the phenomenon
is in English, and then answering questions such as, “What role does the infix play?” or “What
is the goal of using such an infix?”
Approach #2: Socio-political
Socio-political approaches to discourse analysis look beyond the technicalities of
language and instead focus on the influence that language has in social context, and vice versa. One
of the main socio-political approaches is Critical Discourse Analysis, which focuses on power
structures (for example, the power dynamic between a teacher and a student) and how
discourse is influenced by society and culture. Critical Discourse Analysis is born out of
Michel Foucault’s early work on power, which focuses on power structures through the
analysis of normalized power.
Normalized power is ingrained and relatively allusive. It’s what makes us exist within
society (and within the underlying norms of society, as accepted in a specific social context)
and do the things that we need to do. Contrasted to this, a more obvious form of power
is repressive power, which is power that is actively asserted.
Sounds a bit fluffy? Let’s look at an example.
Consider a situation where a teacher threatens a student with detention if they don’t
stop speaking in class. This would be an example of repressive power (i.e. it was actively
asserted).
Normalized power, on the other hand, is what makes us not want to talk in class. It’s the
subtle clues we’re given from our environment that tell us how to behave, and this form of
power is so normal to us that we don’t even realize that our beliefs, desires, and decisions
are being shaped by it.
In the view of Critical Discourse Analysis, language is power and, if we want to
understand power dynamics and structures in society, we must look to language for answers.
In other words, analyzing the use of language can help us understand the social context,
especially the power dynamics.
While the above-mentioned approaches are the two most popular approaches to
discourse analysis, other forms of analysis exist. For example, ethnography-based discourse
analysis and multimodal analysis. Ethnography-based discourse analysis aims to gain
an insider understanding of culture, customs, and habits through participant observation (i.e.
directly observing participants, rather than focusing on pre-existing texts).
On the other hand, multimodal analysis focuses on a variety of texts that are both
verbal and nonverbal (such as a combination of political speeches and written press
releases). So, if you’re considering using discourse analysis, familiarize yourself with the
various approaches available so that you can make a well-informed decision.
How to “do” discourse analysis
Discourse analysis definition
As every study is different, it’s challenging to outline exactly what steps need to be
taken to complete your research. However, the following steps can be used as a guideline if
you choose to adopt discourse analysis for your research.
Step 1: Decide on your discourse analysis approach
The first step of the process is to decide on which approach you will take in terms. For
example, the language in use approach or a socio-political approach such as critical discourse
analysis. To do this, you need to consider your research aims, objectives and research
questions. Of course, this means that you need to have these components clearly defined. If
you’re still a bit uncertain about these, check out our video post covering topic development
here.
While discourse analysis can be exploratory (as in, used to find out about a topic that
hasn’t really been touched on yet), it is still vital to have a set of clearly defined research
questions to guide your analysis. Without these, you may find that you lack direction when
you get to your analysis. Since discourse analysis places such a focus on context, it is also vital
that your research questions are linked to studying language within context.
Based on your research aims, objectives and research questions, you need to assess
which discourse analysis would best suit your needs. Importantly, you need to adopt an approach
that aligns with your study’s purpose. So, think carefully about what you are investigating and
what you want to achieve, and then consider the various options available within discourse
analysis.
It’s vital to determine your discourse analysis approach from the get-go, so that you don’t
waste time randomly analyzing your data without any specific plan.
Step 2: Design your collection method and gather your data
Once you’ve got determined your overarching approach, you can start looking at how to collect
your data. Data in discourse analysis is drawn from different forms of “talk” and “text”, which means that
it can consist of interviews, ethnographies, discussions, case studies, blog posts.
The type of data you collect will largely depend on your research questions (and broader research
aims and objectives). So, when you’re gathering your data, make sure that you keep in mind the “what”,
“who” and “why” of your study, so that you don’t end up with a corpus full of irrelevant data. Discourse
analysis can be very time-consuming, so you want to ensure that you’re not wasting time on information
that doesn’t directly pertain to your research questions.
When considering potential collection methods, you should also consider the practicalities. What
type of data can you access in reality? How many participants do you have access to and how much time
do you have available to collect data and make sense of it? These are important factors, as you’ll run into
problems if your chosen methods are impractical in light of your constraints.
Once you’ve determined your data collection method, you can get to work with the collection.
Step 3: Investigate the context
A key part of discourse analysis is context and understanding meaning in context. For
this reason, it is vital that you thoroughly and systematically investigate the context of your
discourse. Make sure that you can answer (at least the majority) of the following questions:
Discourse analysis definition
• What is the discourse?
• Why does the discourse exist? What is the purpose and what are the aims of the
discourse?
• When did the discourse take place?
• Where did it happen?
• Who participated in the discourse? Who created it and who consumed it?
• What does the discourse say about society in general?
• How is meaning being conveyed in the context of the discourse?
Make sure that you include all aspects of the discourse context in your analysis to
eliminate any confounding factors. For example, are there any social, political, or historical
reasons as to why the discourse would exist as it does? What other factors could contribute
to the existence of the discourse? Discourse can be influenced by many factors, so it is vital
that you take as many of them into account as possible.
Once you’ve investigated the context of your data, you’ll have a much better idea of
what you’re working with, and you’ll be far more familiar with your content. It’s then time to
begin your analysis.
Step 4: Analyze your data
When performing a discourse analysis, you’ll need to look for themes and patterns. To
do this, you’ll start by looking at codes, which are specific topics within your data. You can
find more information about the qualitative data coding process here.
Next, you’ll take these codes and identify themes. Themes are patterns of
language (such as specific words or sentences) that pop up repeatedly in your data, and that
can tell you something about the discourse. For example, if you’re wanting to know about
women’s perspectives of living in a certain area, potential themes may be “safety” or
“convenience”.
In discourse analysis, it is important to reach what is called data saturation. This refers
to when you’ve investigated your topic and analyzed your data to the point where no new
information can be found. To achieve this, you need to work your way through your data set
multiple times, developing greater depth and insight each time. This can be quite time
consuming and even a bit boring at times, but it’s essential.
Once you’ve reached the point of saturation, you should have an almost-complete
analysis and you’re ready to move onto the next step – final review.
Step 5: Review your work
This final step requires you to return to your research questions and compile your
answers to them, based on the analysis. Make sure that you can answer your research
questions thoroughly, and also substantiate your responses with evidence from your data.
Discourse analysis definition
Usually, discourse analysis studies make use of appendices, which are referenced
within your thesis or dissertation. This makes it easier for reviewers or markers to jump
between your analysis (and findings) and your corpus (your evidence) so that it’s easier for
them to assess your work.
When answering your research questions, make you should also revisit your research
aims and objectives, and assess your answers against these. This process will help you zoom
out a little and give you a bigger picture view. With your newfound insights from the analysis,
you may find, for example, that it makes sense to expand the research question set a little to
achieve a more comprehensive view of the topic.
Discourse analysis definition
