Analysing the discourse of the French presidential candidates

Romane Dagain
Extra Newsfeed
Published in
4 min readMay 7, 2017

This is the end of a very disappointing campaign. If you have watched the debate between Marine Le Pen and Emmanuel Macron, the two French presidential candidate, you probably feel the same way than I do now.

More than disappointed, I am mostly frustrated. Frustrated because the candidates’ ideas were inaudible and it resulted in a meaningless conversation up to one point where I was asking myself: “what the hell are they even talking about?!”.

So I decided to take the matter into my own hands and analysed the candidates discourse, using the tidytext package, created by Julia Silge.

I downloaded five speeches for each candidates, made between January and May 2017. I first looked at the frequency in each speech of individual words and then analysed the frequency of bigrams (a unite of two words), which proved to be more meaningful.

Emmanuel Macron, the candidate of “change”:

We can see that Macron presents himself as the candidate of change (remind you of anyone?), with frequent usage of words such as “ambition”, “transformation”, “République Nouvelle” (new republic). He also adopts a modern position by using expressions such as “French Tech” very often.

Most frequent words in Macron’s speech
Most frequent words used by Macron, per speech

There is a shift in the tone of his last speech, the 1st of May, after the first round of the election. He launches a clear attack against the Front National by using tougher words such as “colère” (anger), “mensonge” (lie), and “extrême” (extremism).

Most frequent expressions in Macron’s speeches

Marine Le Pen, “France Reunited”:

Interestingly, when looking at the frequency of individual words in Le Pen’s speeches, I noticed that the texts were dominated by words with positive connotations: “apaisée” (serene), “vérité” (truth), “diversité” (diversity). This can be interpreted as an attempt of Le Pen’s to soothe her discourse and revamp the Front National into a more traditional political party.

Most frequent words in Le Pen’s speeches
Most frequent words used by Le Pen, per speech

When looking at word associations, we can see that her discourse on immigration is presented as an economic argument first (with expressions such as “puissances d’argent” (money power), “petites retraites” (small pensions), “mondialisme économique”(economic globalism), “mondialisation” (globalisation)). To illustrate her argument, it is important to note that she uses heavily large figures (“75000”, “50 milliards”, “1.2 million”) as a mean of legitimisation.

Most frequent expressions in Le Pen’s speeches

This economic argument is closely intertwined with more ideological terms (“sentiment national”, “sentiment d’appartenance” (feeling of belonging)) and is overshadowed by the theme of terrorism, a key element of Le Pen’s campaign: “terrorisme islamiste” (islamist terrorism), “islamisme radical” (radical islamism).

Two opposite candidates?

When reading the press, we hear over and over again how different these two candidates are: one is a liberal pro-Europe, the other an Eurosceptic nationalist.

They may defend different views and policies, however their discourse have interesting similitudes. They both present themselves as anti-establishment by referring heavily to the “système” and they both put emphasis on the theme of public safety.

Finally, I decided to do a simple comparative word-cloud. We can easily guess major themes of Le Pen’s discourse by simply looking at the word-cloud, however it is more difficult to find a main angle in Macron’s speeches, something that has been often criticised during the campaign.

This is obviously a very superficial analysis, as I only considered selected speeches over the course of the last 5 months. If you are more interested in that topic, I recommend you to check out Cécile Alduy and Stéphane Wahnich’s work, who analysed the evolution of the discourse of the Front National between 1987 and 2013.

--

--

Romane Dagain
Extra Newsfeed

R enthusiast who likes to tell stories using data and technology.