Referencing Lemma before introducing it

Question

I am always confused about this: When I am writing a paper I would like to first provide the big picture and then address some detail issues. On the other hand, I think it's common knowledge that one should not reference something which will be introduced only after the reference.

For me these tow statements are conflicting each other.

My concrete problem is that I would like to first state the Theorem, then prove it using a some technical lemmas which are introduced later in the manuscript.

Since the proofs are not too long, I do not like to put them in a separate proof section.

How do you usually deal with this top-down vs. bottom-up dilemma?

Answer 1

I've seen this handled several different ways:

1. Just go ahead and cite the lemmas from the future sections, possibly with a sentence or two of warning beforehand.
2. Put the lemmas in section $n$ and the theorem in section $n+1$, and at the beginning of section $n$ say something like "The reader is advised to skip over this section and return later."
3. State the lemma twice, the first time without proof and the second time with. (I'll commonly see this in introductions where someone will put a copy of the statements of the various pieces and summarize the way they fit together.)

I guess it depends to some extent on how applicable the lemmas are. Could they be of use for any other problem, or are they inextricably linked to your one theorem?