If so, how would that workflow look like?

I prefer authoritative sources for such a workflow/diagram, but you can make your own if you feel like that. I like yuml for that, but feel free to use whatever you prefer.

PD: examples: 1, 2, 3, 4, 5. As I said, I'm searching for authoritative sources.

  • 1
    How is this about philosophy? It seems like you are asking about the practices of working scientists, right? Wouldn't this be better suited for one of the science SEs where you might find a larger number of working scientists who can explain to you what their workflow is? – Dennis Aug 3 '13 at 18:22
  • @Dennis this is epistemology and philosophy of science. Metaphorically we can say some guys made a car and some other guys drive it in the streets. If you want to know the internals of the car, you ask the first group of guys. – Trylks Aug 3 '13 at 18:23
  • 1
    I saw the tags, but your metaphor just sort of confuses me more. What aspect of science are you asking for a workflow for? I thought you were asking for a workflow describing scientific practice, in which case my previous comment still stands. It seems that maybe you aren't, though? – Dennis Aug 3 '13 at 18:43
  • @Dennis "Science is the systematic effort to acquire knowledge in the form of testable explanations and predictions about phenomena in the world." (source) That "systematic effort" sure can be explained in more detail with a workflow, right? – Trylks Aug 3 '13 at 18:56
  • 2
    You might be able to find information on scientific workflow systems at the scientific workflow systems wiki. ;) – David H Aug 15 '13 at 15:04

Think of science more like a cycle. Or rather a linear workflow is never ending and wraps in on itself. In order there is:

  • observation - recording what seems to occur
  • description - labeling and naming the things observed
  • hypothesis making - theorizing connections or implications among the definitions
  • hypothesis testing - experimentation, trying out the implications
  • reassessment - when the hypotheses don't completely match up with the observations, go back and fix the definitions or theories.

This is neither normative nor descriptive of the process of science, just an explanation of one possible thread of activity where one task follows another logically. But of course such a linear ordering of these stages never happen in practice so orderly. Just observation depends on your expectations which are driven by theories. And this may go in a big cycle or in many small cycles simultaneously or overlapping in parallel.


  • Yes I have a similar idea and that's similar to what we can see in the examples I added. However I'm not certain about how much consensus exists on these descriptions. Usually there is a reasonable consensus in the big picture that breaks when getting closer to the details. – Trylks Aug 15 '13 at 12:57
  • I'm not reading from a reference; I'm sure there are a number of treatments that give a linear process (which of course is a gross simplification because people never come to a situation with no definitions or observations or hypotheses. – Mitch Aug 15 '13 at 12:59
  • Even if science was defined with a linear workflow it would be a workflow that would be re-started many times, every time someone makes science. So the linearity doesn't change much IMHO. However the reassessment could be considered the first step and not the last one, because it is more usual to search for limitations in the state of the art (made by other people) than the work done by yourself. – Trylks Aug 15 '13 at 13:04

Can science be described with a workflow?

No. A workflow for doing Science does exists, and the most well known is the "Scientific Method"; but "Science" itself is more about the idea rather than the "Scientific Method" which is the way.

The idea of Science itself is based upon empiricism (observation trumps any theories), falsification (truth are determined by elimination of hypothesis from contradicting evidence instead of confirming evidence), objectivity (the accuracy of observations can be tested independently from the individual scientist who first reports it), and probably a few others I can't quite recall at the moment.

The Scientific Method encodes these ideas into a general algorithm that, if followed correctly, should lead you to truth through a process that conforms to the idea of Science.

  • That's very interesting. I'm interested in what I understand to be two different things (1) the method (in other answer it was referred to as scientody) and (2) in those ideas you mention and those you can't recall now. Would it make sense to open a new question for those ideas? How would you ask for them? (I'm afraid my background knowledge is not enough to pose the question properly). Thank you. – Trylks Aug 18 '13 at 21:31

Not the answer you're looking for? Browse other questions tagged or ask your own question.