I suspect that for many of my first-year Biology students, the sheer weight of new terms they come across is perhaps the most daunting thing about the course. In some ways learning biology is rather like learning a new language - with several thousand new words swamping the page (& the brain) over the course of a 3-year degree.
But there's more than just the new words - there's the meaning of the words to come to terms with. This is the focus of Helen Quinn's 2007 paper, Belief and knowledge - a plea about language. There are many words where their meaning to a scientist may be quite different from what they mean to a lay person. Quinn feels, & I agree, that some words 'are the root of considerable public misunderstanding about science: belief, hypothesis, theory and knowledge.'
'Belief' isn't really a word that sits well with science. It has a couple of meanings in everyday speech. As Quinn says, it can be 'an article of faith' ie religous belief. Or - conversely - in the phrase 'I believe he is coming at 5pm', you get the meaning 'but I'm not really sure.' So how are we to take those news stories that begin 'Scientists believe'? A statement like 'most biologists believe in evolution' could be used to claim that evolution is as much faith-based as organised religion. (I tell my students that I don't 'believe' in evolution, but accept it as the best available current explanation for life's diversity. This can engender some interesting discussions...)
But what the statement 'most scientists believe' means - to scientists - is that the majority of scientists are in agreement that the weight of evidence favours a particular interpretation, & that for now there's no evidence to contradict that interpretation. Quinn suggests we should say 'scientific evidence supports the conclusion that...' I like this - it leaves open the possibility that this conclusion could change, if sufficient evidence to the contrary comes to light. Which is a much better reflection of the nature of science: that its conclusions are subject to change if the evidence demands it. Unfortunately there tends to be a perception that scientific 'facts' don't change. (Also unfortunate is the fact that if scientists do change their interpretation of the data, they're accused of not really knowing what they're talking about by those who don't understand how science operates. Sometimes I think we just can't win!) Like Quinn, I feel that as scientists we shouldn't be using the 'b' word - it gives the appearance that science is 'just another belief system.'
'Theory' is another word that means different things to different people. 'I've got a theory about that' really means, 'I've got a hunch or an idea, a guess.' But to scientists 'theory' means a well-established explanation for a large body of data: the theories of relativity, plate tectonics, evolution... These are definitely not guesses (nor are they belief systems!), but comprehensive explanations that have strong predictive power & have been tested time & time again. They are also incomplete, but that again is the nature of science. Scientific theories may well be modified if new evidence comes to hand: Newton's laws are an example. (Quinn notes that Newton's laws still hold, under certain well-defined condtions - they're weren't just thrown out when Einstein & special relativity came on the scene.)
To finish, it's worth repeating Quinn's description of how scientific theories are developed, because this is a valuable description of how science operates and what sets it apart from 'other ways of knowing':
When we seek to extend and revise our hypothetical frameworks, we make hypotheses, build models, and construct untested, alternate, extended theories. These last must incorporate all the well-established elements of prior theories. Experiment not only tests the new hypotheses; any unexplained result both requires and constrains new speculative theory building - new hypotheses. Models... play an important role here. They allow us to investigate and formulate the predictions and tests of our theory in complex situations. Our theories are informed guesses, incorporating much that we know. They may or may not pan out, but they are motivated by some aspects or puzzles in the existing data and theory. We actively look for contradictions.
H.Quinn (2007) Belief and knowledge - a plea about language. Physics Today January 2007: 8-9