Actually, there is such a requirement in the practice of science. Particularly when there is no sound argument against the currently accepted theory.
That is ludicrous and illogical.
There is no NEED for a theory.
A theory stands on its own.
Lack of another theory is not supporting evidence for a theory.
Take observed UFO's. There is no sound theory to explain them but that does not prove they are flown by space aliens (which is a theory)
Your statement is similar to saying: one must prove who actually did a crime before one can be found innocent.
Actually, there is such a requirement in the practice of science. Particularly when there is no sound argument against the currently accepted theory.
It sounds like you are overstating your case somewhat there, JS. One can question a theory until the cows come home without an alternative theory, under certain conditions, and still be practicing science.
Some of these conditions might include:
One has new experiemental evidence which appears to contradict the theory
The theory being questioned is new, and one is questioning the mathematics, the predictions, or the design of the experiments which support the theory
"Less-filling / tastes great" debates such as nature vs. nurture, or low-carb diet (Atkins) vs. low fat (Ormish)
And of course one is always free to look for logical flaws or faulty predictions, even if one has no replacement.
Despite some posters' intellectual vanity, it really is acceptable to say "I don't know" or "I haven't studied that yet" or "Gee, how does that model work in regard to XXX?"
Full Disclosure: It may be true that none of these examples apply to the typical posting by cre-o's here. But there is stil such a thing as refining a model; or such a thing as poking holes in a model, in order to alert people it's time to work on a better one--even if you haven't worked out the improvements yourself.