Here's my attempt at a definition, not that I think it's necessary to have a clearly defined definition - we're intelligent beings and able to deal with complex, abstract concepts in any number of ways. We select tools for the task, if they're not up to the job, we either improve them or discard them and use other, more appropriate tools. I find myself repeating phrases from other pages...

Contrast that to the approach at school. At school it's acceptable to "not be good at maths" or "I can't do science". A reality crackers approach is "I want to achieve this. This is how I'm going to start. If that doesn't work, I'll plan again and approach it differently. I'll finally achieve it because (motivation goes here). I might have to leave it for a while and return refreshed". Notice the difference?.

Reality cracking is a process of analysis and investigation for the purpose of making better, more informed decisions. There you are. Sounds a bit pompous. There's a bit more to it than that. In reality cracking we've already rejected the obvious explanation as inadequate. We need more substance, more depth, we know that the explanation or information given is inadequate, incomplete and wanting.

So how are we going to achieve that? I'll tell you. We're going to achieve better, more informed decisions by taking control of the process and doing it better. I tend to analyse all the concepts I use quite deeply and then try to follow the path between them. One of my favourite reality reversing tools is multiple causality. This is how it works. We refuse to accept that there is only one simple cause to an event, we insist that there are more. So for example, if I find my car doesn't start, I can't just say that the battery's a bit low because that's far to simplistic and doesn't give us the depth of understand that we as reality crackers are searching for. Let's try again. The car doesn't start because the battery's low and that's because the fan belt's loose, and because it hasn't been started for 2 weeks, and it's very cold, and maybe I should repair the car and because the car didn't start, I'm not able to go to work today. Yipee! On the other hand, if the car had started I'd have to go to work. Maybe I should get a different job.

If you do choose to use multiple causal attribution, you need to consider it's effects on the other concepts you use. So, say you get into a physical fight with someone, you're not able to attribute the cause of the fight totally to the other person because that would be simply causal. So, if you can't blame only the other person for causing the fight, who can you blame?* ~ Did your own presence contribute to your involvement in that fight? In other words, would you have been involved in that fight if you weren't there at that time? In one simple step, you've reduced or eliminated the 'self-serving bias' in attribution and you're making better decisions.

One topic that seems to raise it's ugly little head in other reality essays is the appeal to Empirical or 'scientific' knowledge. This needs to be addressed. Scientific knowledge has no more validity than any other variety of knowledge. That started some arguments. This is how I understand the scientific method. Think of a theory (hypothesis). The theory stands until it's disproved. If you prefer a competing theory you can swap over to that one as long as it explains the facts and has not been proved wrong. If it does get proved wrong, you can either swap over again or amend the theory. Theories are very limited in what they can explain.

This is how the 'scientific method' (how to do experiments) works. The experimental environment is controlled in order to (hopefully) exclude 'extraneous' variables that might interfere with the experiment. An 'independent' variable is manipulated under the experimenter's control. The independent variable's affect on the 'dependent' variable is measured. Statistics are applied to the result. Results are considered 'statistically significant' if there is a less than one in twenty chance of the results happening by chance.

In assessing 'external validity' (is it possible and correct to apply the results to the wider environment outside the controlled experimental conditions?) you should consider the experiment design. The design reflects the theory on which it is based. I think of the experiment as being within the theory. Does it measure what it claims to be measuring? Can it be applied outside the experimental environment?

I see three main problems with the scientific method. The first is 'experimenter effect' where the dependent variable is affected by the experimenter. This happens to atomic particles as well as when experimenting with people. With people, they tend to do what they think the experimenter wants them to do (also happens with questionnaires). The experimenter effect is considered coincidental and unintended. You can actually learn how to reduce experimenter effect. The second problem is that of external validity which also has to do with the third problem... ... that it seems to deal (fairly poorly) only in simple single causal attribution...

Concerning external validity, experiments with people are conducted on groups. What you're measuring is general tendencies within groups of individuals. I think that it's quite wrong to then apply that knowledge to individuals. Why? Because everyone's different and the variables are measured in isolation. By the very nature of science, you only have a very partial picture. Yes scientific theories do provide insight, but that's all. It's good enough for the breadheads' purposes, but not good enough for ours. We can make far better use of scientific theories and results by understanding their limitations and applying them appropriately.