Originally Posted by Shaggy Hiker
Your example has one fatal flaw: Churchill had the advantage of hindsight. He warned about the growing theat of the Nazi party....and history proved him right. Aside from being argument by anecdote, and the problem that you have a case of n=1, that argument is bizarre. How many people warn us about various threats? Some of them are right!!! Gee, that's a surprise. Does this prove anything about the general case? No. Does it give us any information at all about the general case? No. So, of what value is it? Well, it allows people to make arguments like that one.
However, consider, if you will, the consequence of that action. Assume that all people saw the world with the clarity you assert. Radical Islam is not attacking us just because they are bored. They weren't sitting around reading comics one day and decide "Let's go attack somebody." Instead, they had a plan. Frankly, it doesn't matter what the plan was. It doesn't matter if they thought things through logically, if they threw dice on the subject, or if they were completely insane.
What matters is that they looked at the world, and whether rational, crazy, or random, they decided that group x was a threat to their beliefs, just like we can reasonably say that they are a threat to our beliefs.
Some people say that "Appeasement doesn't work" based on Chamberlain. As long as you can assume that you are always right, then this policy is valid, because everybody who opposes you is wrong by definition. However, the bulk of us realize that we have just an oppinion, which is not divine in nature. If you approach the question from that vantage, then saying "Appeasment doesn't work" has to hold true not just for you, but for anybody who disagrees with you. There is no scenario short of divinity in which you can rationally assert "I must not appease, but my opponents must all appease". If instead you assert "I must not appease, and my opponents must also not appease", then you must go to war over EVERY disagreement, because only capitulation is an alternative.
Now, if you've followed that rambling explanation so far, you may have noticed that this is closely linked to the game theory problem known as "The Prisoner's Dilemma". Cooperate for a shared minor good, or hold out. The payoff is somewhat inverted over the classical case, but the various options have the same relative weights. You lose most if you appease, and your opponent does not (Nazi Germany). You gain most if you don't appease and your opponent does. You both gain at a lower level if you both appease (negotiate).
This puzzle has been tested exhaustively. Douglas Hofstader (might have that spelled wrong, best known for Godel, Escher, Bach), or Dennett, or both, described a contest where computer programs were pitted against each other to see which strategy won. The "Do Not Appease" strategy was submitted as one of the contestants (by somebody who shares my last name, though I don't know him). That strategy did fairly well, but was not the winner. The winner was a strategy known as Tit for Tat. You'd have to hunt down the book for a more thorough review. I also heard that a later test used more sophisticated strategies, and Tit for Tat was beaten by Tit for Two Tats.
Thus, I disagree with your argument. You can only know when talking is no longer viable in hindsight, and relying on "Appeasement Never Works" is not supported.