[The_Sword_of_Groo]

Sure you could, If, as you mentioned, you had a control group or enough data to compare. I am speaking in general, conceptual terms. "The Fall of Rome"...ok, whats your hypothesis? "Did Rome Fall" ? Yes. next question. Like I said, garbage in, garbage out. Its a poorly worded topic. Would you have to be more specific? Of course. However, you can easily study the causes of decline in civilizations using any number of criteria and a control groups. Examples that have been studied comparing greeks and romans to modern day civilizations: Rates of disease and methods of sewage treatment; any number of medical treatments, global warming ( barf), immigration patterns, etc, ad nauseum.
Additionally, in an ad hoc study, all events are predictable. And lets be honest...if you didnt think you could identify indicators that an empire would fall (unpopular leadership,unchecked immigration, over regulation, disproportionate tax rates) you probably wouldnt be posting on FR.
My point is, in theory, any topic that can be studied can be studied using the scientific method provided the search area is large enough or the time frame is large enough to provide appropriate data.

[allmendream]

The Fall of Rome is not reproducible or conducive to controlled experiments; where with ONE Rome you hold everything constant - and with a DIFFERENT Rome you change one variable.

You can study literature as rigorously as you wish - it will not be science. Science is not just rigor. It is a method for gaining reliable information utilizing experimentation and observation to develop useful and predictive models about the natural world.