The AfD Monitor reads the party's own YouTube channels. It looks at what the party posts, and at what people write in the comments. The same rules sort both, so you can set the party's words beside its audience's, on the same videos.
…
A computer sorted every comment and video on this site, using one written set of rules. No person chose where any comment went. So the sorting can be wrong, but it is wrong the same way every time. It cannot lean toward an answer someone wanted.
Each number comes with a range, drawn on the chart. The true figure is probably inside it.
Treat the sorting as a careful first reading, not a final verdict. Every rule is in the codebook tab, and every video is listed in the corpus tab.
Video titles and what is actually said in them, taken from the AfD's own channels.
…Comments on those same videos. Replies are skipped, because a reply only makes sense next to the comment it answers.
…The site compares the party only with its own audience. It does not compare the AfD with any other party.
The charts above combine every week into one figure. Below, each week is shown on its own. Read each week next to the videos the party posted that week. People are replying to those videos, so the numbers move partly because the subject changed.
On the party side we count videos, not sentences, because ten sentences taken from one video are really one source. That is why the ranges on the party's bars are so much wider than the audience's: 27 videos is a small number, and the ranges show that honestly.
Every video and comment behind these figures is listed, with links, in The corpus.
Every figure on this site was computed from the videos below and the comments underneath them. Each is named and links to the original, so any number on this site can be traced back to the material it came from.
| Video | Week | Comments analysed | Transcript pieces | Views |
|---|---|---|---|---|
| Loading… | ||||
The AfD Monitor started with a simple question: what do the AfD and its supporters actually say, when you look closely at their own words?
Most public debate about the Alternative für Deutschland (AfD) asks one question: is the party a threat to German democracy? At the same time, the party keeps winning votes. So two stories compete. One sees the AfD and its supporters as extreme and dangerous. The other sees a party speaking for concerns that the other parties have failed to address.
This project starts somewhere else. It does not begin by judging what the AfD is. It asks what the party actually says, and how its audience responds.
The AfD Monitor reads the AfD's own YouTube videos, and the comments people leave under them. It asks four questions:
My research on political language, threats and group identity shaped these questions. But a computer does the sorting, using one written set of rules for every comment. No comment is picked out, or sorted by hand, to support a point. The computer can make mistakes. The aim is to use the same method every time, and to show where it falls short.
So this project does not try to prove that the AfD is a danger to democracy. Nor does it try to prove that the party is a fair answer to Germany's problems. It tries to show what the party and its audience actually say, before anyone decides what it means.
The figures, the rules and the list of videos are all public, so anyone can check them and challenge them. The comments themselves are not published, to protect the people who wrote them. We never keep a commenter's name or account either. Each one is stored only as a scrambled code that cannot be turned back into a name, so nothing on this site can be traced to a person. The categories are careful judgements, not proven facts. Read them as a first look at a large amount of political speech.
The goal is simple: to understand political movements through what they say themselves, and what their audiences say back, not only through what others say about them.
I am Talip Alkhayer, a political scientist specialising in political rhetoric, threat framing, group identity and computational approaches to the analysis of political discourse. My academic research combines theories from political science and social psychology with methods from natural language processing and machine learning to examine how political actors construct threats, identities and narratives.
I developed the AfD Monitor as an extension of my broader research interest: using computational and AI-assisted methods to examine contemporary political discourse at a scale that would be difficult to analyse manually.
I am interested not only in what political actors say, but in the patterns that emerge when thousands of individual pieces of discourse are analysed systematically.
This is an ongoing research project, and it is meant to be questioned.
Have you spotted a problem with the method? Do you disagree with how something was sorted, or have an idea to improve the codebook? I would like to hear from you.
Message me on LinkedInEvery figure on this site comes from a computer reading text and sorting
it into groups. Below are the exact rules it was given, copied straight from its
files. The short names in code, like present_grievance, are the labels
the charts show in plain words.