Some days ago, in discussing "intelligent design, I accused proponents of this philosophy of not knowing what the word "theory" meant. Upon further reflection, I have come to the conclusion that at least some of these proponents are well aware of the fact that the word "theory" has two quite different meanings ― and are more than happy to confuse the two meanings in order to confuse their listeners and push their ... well, let's call it a "theory" about the universe and its origins.
As I suggested, in ordinary parlance the word "theory" means pretty much the same as "unsupported supposition". As an example: I hear a gunshot from the street behind our house. I explain to my wife: "My theory is that the guy who lives around the corner ― I've never much cared for him; he has a shifty look to him, his eyes are too close together ― just put a bullet into his son, who, I'm pretty sure, has been dealing drugs." Later, of course, it becomes known that a completely different guy in a completely different house was cleaning his .38 when he accidentally put a bullet into the wall. (My "theory" that the son of the first guy has been dealing drugs may or may not be true, but is now irrelevant, since he just received a full scholarship to Stanford ...)
I've just used the word theory in the colloquial sense: an unsupported explanation for an observable phenomenon. Same as supposition.
To a scientist, theory has quite a different meaning. The scientist, having heard the gunshot, might hypothesize that the disliked neighbor had shot his son, but he would not dignify the hypothesis with the name theory. Instead, he would stroll around the corner to the first man's house, knock on the door, and ask: "Did you just have some problem with your gun?" To which the man might well respond: "Not I, I don't even own a gun. But I have a lawnmower that needs fixing; do you have any experience with those?" The scientist, having disclaimed any understanding of the esoterica of lawnmower motors, would then visit each house on the block until he came to the one where the owner, answering the door, says: "Yeah, I was cleaning my .38 and it went off. Now I have a bullet hole in that wall over there." At which point, the scientist would return to his wife and explain his theory (which remains a theory, since he wasn't actually present when the gun went off, and the guy could, after all, be lying): "I believe that Joe Rodriguez, back behind us, was cleaning his gun and it went off." (1)
I hope this clarifies the difference between a colloquial theory and a scientific one. The colloquial theory is nothing more than rank supposition (though in many cases it may, by simple coincidence, prove to be right); the scientific theory depends on (a) explaining an observed phenomenon, and (b) not going against any other observed facts. It should also be usable to predict future phenomena related to the original phenomenon (the scientist's prediction: Joe will not be hauled off by the police to stand trial for filicide).
Unfortunately, when a proponent of "intelligent design" says, of evolution, "it's only a theory", he is taking advantage of the fact that many, perhaps most, people don't understand the difference between these two meanings of the word "theory". Given that at least some such proponents claim scientific training, they themselves have to be aware of this difference ― which means, to me, that they are intentionally trying to muddy the waters.
Well, that's my theory, anyway.
(1) Another possibility ― given that Joe could have been lying ― is that Joe's wife, an hour earlier, had asked him why he even had a gun, given that he couldn't hit the broad side of a barn with it ― even if he was in the barn. And Joe was simply trying to prove his wife wrong, by trying to hit the wall of his house from the inside. When a scientist has two possible hypotheses that explain all the facts ― a cleaning accident vs. an intentional experiment ― he often depends on something called "(William of) Occam's Razor", aka the "Law of Least Effort", to determine that the hypothesis which explains a phenomenon in the least baroque fashion is the likely one. This is, however, a philosophical idea, not a scientific one, and it can prove to be wrong. Nevertheless, a scientist would, in the absence of further evidence (eyewitness testimony), probably choose to accept the "cleaning accident" rather than the "intentional experiment" explanation. And, of course, in the absence of a body, he would never theorize that Joe shot his son.