An argument is the smallest meaningful component within an argument map. It is an assertion that has been made (e.g., by a person or an organization). Some arguments are written by Controvis users to summarise an interesting assertion or group of assertions. Others are direct-quotes from a source. Quotes contain sources, with links (where possible) to source webpages and metadata pages containing additional contextual information about speakers and authors.

Argument Visualization

Arguments can appear in two formats on the map:

• Pinned (expanded view) - Displays the full text of the argument, allowing you to read the content directly.
• Unpinned (collapsed view) - Represented as a small box with tags, which helps simplify complex maps while preserving all arguments.

Connections between arguments show their logical relationship. There are three main types of relationships in Controvis argument maps:

1. Support Relationships

In a support relationship, one argument provides evidence or reasoning that strengthens another argument. For example, if Bob says that he has been refereeing for 20 years without any problems, this supports the claim that he can be trusted to referee the match.

When an argument is supported and not successfully attacked, it becomes acceptable in the debate.

2. Attack Relationships

Attack relationships indicate that one argument provides evidence or reasoning that weakens another argument. There are three important types of attacks:

Rebutting

A direct attack on another argument. In the example, Alice's argument directly rebuts the main claim that Bob can be trusted to referee the match. If there was just the main claim, Alice's attack, and Bob's support, then the main claim would be deemed unacceptable since it is attacked but not successfully defended.

Undermining

An attack on an argument that supports or attacks another argument. Bob's counter-argument that he made the comment eight years ago undermines Alice's attack. Given just these three arguments, the main claim becomes acceptable again because it is now defended from all rebuttals.

Key insight: As arguments are added or removed, other arguments can become acceptable or unacceptable—this dynamic nature is what makes argument maps more powerful than simpler representations.

Countering

A counter-attack defends an argument by attacking the argument that was attacking it. These counter-attacks can continue in chains. In the example, Alice provides a further counter-attack, claiming Bob made the same statement last year as well. With this new information, the main claim becomes unacceptable again!

Undercutting (not supported yet)

For completeness, a fourth type of attack exists called an undercut, which is not currently represented in Controvis. An undercut attacks the relationship between two arguments rather than an argument itself. It suggests that while argument A might be correct, it doesn't actually attack or support argument B in the way claimed.

3. Citation Relationships

Citations indicate that one argument depends on another by referencing it as evidence. Unlike support or attack relationships, if a cited argument is attacked, it doesn't necessarily affect the argument citing it.

Citation relationships are particularly useful for mapping arguments about people's values, ideologies, or positions while allowing those underlying arguments to be explored separately as part of the broader map.

Controvis documents arguments and their relationships to analyze them, not to endorse them. While some arguments may be potentially harmful at an individual or societal level, we permit them only when:

• Including them reduces rather than increases potential harm
• They provide valuable context for understanding the broader debate
• Counter-arguments are included that challenge harmful positions

For example, we may include claims that dehumanize certain groups if doing so helps expose problematic ideologies and values, while always providing counter-arguments that explain why such claims are harmful and incorrect.

We focus on creating and maintaining argument maps that are:

• Complex - Covering topics with multiple valid perspectives where visualization adds clarity
• Interesting - Exploring topics worth understanding in depth
• Valuable - Comprising high-quality arguments with meaningful relationships between them