Reenigne blog

I was arguing (in a friendly way) with some people the other day and it got me thinking about an idea for a new website. Technically it is a kind of "conflict resolution" website, but in practical terms it is a forum for people to argue in without anybody getting angry or repeating themselves and without the argument devolving into what the definition of "is" is.

It occurs to me that this idea has been brewing in my head for some time - back when I was living at home, one of my brothers used to argue with my parents quite often all the time and I used to try to moderate the arguments. It was frustrating to try to get each party to actually listen to what the other was saying. Also, because the refutation of any given statement might involve several other statements (more than one of which might need to be refuted) it can get difficult to keep track and some threads of conversation might get forgotten.

Here is how the site would work. User A posts a controversial statement - say (for argument's sake):

• Statement 1: "Abortion should be legal"

This statement becomes a page on the site. User B, happening upon this page, is given the option of expressing an opinion on this statement. The options are "I agree", "I disagree" and (the default) "I neither agree nor disagree". On each statement's page, the site would show the proportion of people who have expressed an opinion on the statement and the proportion of those who agree with it.

Suppose user B disagrees with statement 1. They then have the opportunity to make a set of arguments and statements in support their opinion. In this case, such statements might be:

• Statement 2: "Abortion is murder"
• Statement 3: "Murder should not be legal"
• Statement 4: "Statements 2 and 3, when taken together, imply that statement 1 is false"

User A might agree with statements 3 and 4 (and can press buttons on the pages for those statements saying so) but disagree with statement 2. User A's arguments against statement 2 might take the form:

• Statement 5: "Murder is the deliberate premeditated ending of a human life"
• Statement 6: "An unborn fetus less than 24 weeks into the term of pregnancy is not a human life"
• Statement 7: "Statements 5 and 6, when taken together, imply that statement 2 is false"

User A might also make another statement in support of statement 1:

• Statement 8: "If abortion was illegal, it would create a black market for abortions which could be dangerous for women who are in desperate situations"
• Statement 9: "Statement 8 implies statement 1"

User B might then agree with statements 5, 7 and 8 but disagree with statements 6 and 9 and give their reasoning. And so on.

There are two sorts of statements here - "monolithic" statements like statements 1, 2, 3, 5, 6 and 8 and "connecting" statements like statements 4, 7 and 9. The site uses connecting statements to create trees of deduction for a particular user - figuring out why people believe what they believe.

The site would keep track of unrefuted statements, track back through the deduction trees and show, for each statement, which side (if any) is "winning" the argument.

In the course of reading statements from other users, a user might change their mind about some statement. Suppose user C agrees with statements 3 and 4 but disagrees with statements 1 and 2, but is then persuaded to change their mind about statement 2. The site uses the trees of deduction to find contradictions in a user's opinions. The user will then be given the opportunity to choose which other statements they want to change their mind on to resolve the conflict. In this case, the site would say "Your opinions on statements 1, 2, 3 and 4 conflict. Which one(s) do you want to change your mind on?"

Another possibility is that two users might disagree on some experimentally verifiable fact. The site would keep track of such statements and, given a disagreement, would be able to propose an experiment which would resolve the disagreement one way or another. Once this experiment has been done, the result (with suitable references) can be posted in another statement and can be used to convince people of things or can be disagreed with (if people believe the experimental method used was flawed).

Yet another possibility is that two users agree to disagree on a particular statement, and make a connecting statement saying ("statement x is neither provable nor disprovable"). That is fine too - the site only records people's opinions on various statements, it does not pretend to be fount of universal knowledge.

Further refinements which could be made:

• The site could keep track of definitions of particular words, so that users have a common vocabulary to communicate with (e.g. the word "murder" in statement 3 might be a hotlink leading to the page for statement 5). A statement about murder in a different context might link to a different definition.
• Users could associate themselves with particular groups such as "liberal", "conservative", "libertarian" - this would not change any of the arguments but would allow the site to be able to generate interesting (although possibly obvious) statistics such as "90% of conservatives agree with statement 2". The site could even try to categorize users based on their opinions.

Of course, some users might disagree with the logic that the site uses to deduce things. For example, a user might disagree with "if A implies B and B implies C then A implies C" on finding some of their long-held beliefs in contradiction. Such users would be encouraged to set up their own site with their own system of logic (you can't even begin to have an argument if you don't agree on some basic axioms that you can use to deduce things). Of course, a similar website which uses a form of logic that isn't very useful won't be well-frequented!

This entry was posted on Wednesday, August 31st, 2005 at 8:18 pm and is filed under computer. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.