How can Business Analysts prove anything when ultimately all their information comes from other people's opinions? Science can prove stuff - why not us?

Let's contrast how science and Business Analysis proceed:

Science has an objective to uncover more new knowledge. Science proceeds from observations on reality - these are then explained using 'hypotheses' and tested experimentally: if the hypothesis is true/false and we do such-and-such then this should happen. Those hypotheses that survive testing become theories. Theories are what pass for 'facts' in science on the basis that even though you have proved experimentally 10,000,000 every time you drop the hammer it falls, the 10,000,001st time it may not invalidating the theory of gravity that explained the falling hammer.

Business Analysis has an objective to uncover more new change requirements: Our 'facts'. But Business Analysis does not have reality as a starting point like science does.

Luckily (maybe) there is a form of reality that can be defined for Business Analysis but it can change - it is not as fixed as the reality that science has the luxury of: our BA reality is the decisions made by a group of people who have the recognised authority (formally or not) to sanction our project to proceed or kill it. Crucially, they must agree a joint definition of what needs to change and by how much in order for the project to be considered successful (smart Objectives).

These 'killer stakeholders' are not always who you would expect: sure, the budget holder is there. But what about the guys in charge of the IT standards and procedures? Try implementing without their agreement and you will rapidly experience the effect of killer stakeholders. Their objective: this project must maintain compliance with IT standards and procedures.

Once we have these smart objectives we can proceed to define what has to change in order for these objectives to be met (change requirements). Once we have that we can define how the changes will be made (design a solution) that will deliver these objectives. Then we can build, test and roll out.

Almost all change projects will have killer stakeholders we may not have expected: who is accountable for compliance with the Data Protection Act? Who is responsible for validating the business case? What about John Smith the main user because if he won't use it no-one will? And so on.

BA reality is built from people (mostly) who can leave their jobs, get promoted, and just change their minds! All our analysis is built using logic that rests on the premise that the smart objectives are right - and that premise depends on the agreement of all the killer stakeholders.

Then we can start to build up all the fundamental components of Business Analysis in to a chain of reasoning that can be demonstrated on demand to prove that our analysis is correct.

And scientists thinks they have it tough?!?!?!?