(Step one was here, step two was here.)
Perhaps I got a little carried away at the end of step two, when I wrote that the next step would be to formulate the opening argument, the cross-examination and so forth.
Before you can do that, you need to put some serious thought into your theory of the case: What are you trying to convince the jury of? Are you conceding it happened, but it wasn't your client that did it? Are you saying it never happened? Are you convincing the jury that someone is lying or mistaken? And if so, who?
I think the biggest mess a defense attorney can get themselves into is not knowing their theory of the case.
If your goal is to prove, for example, that the victim lied and she wasn't robbed, your line of questions will be different if your goal is prove that the victim was robbed but she picked the wrong guy out of the lineup. (You can simplify this as the difference between "whodunit" and "what happened.")
Once you have a theory of the case, you want to imprint that into your mind and let it control every question you ask.
In the bank robbery case that I used as an earlier example, I probably can't deny that the bank was robbed. So my defense probably has to be that the bank was robbed, but my client didn't do. The details would probably depend on how many people witnessed it, whether my client was wearing some kind of disguise, and whether my client made any confession.
Sometimes it's hard to have a theory of the case. If your client was seen in broad daylight by ten witnesses, all of whom are credible, and he was caught on video which is clear and not grainy, and he made a full confession, and he was caught with the stolen money in a bag with dollar signs painted on it, just like in old movies... you may be walking into your trial without much of a defense, just hoping that something good will break your way. Sometimes you're going to trial on only the unlikely possibility that the prosecutor forgets to ask a key question, the witness suddenly forgets something while testifying and the video won't play. It happens, but keep in mind that your defense is pretty open ended (you're just hoping for anything good to happen) and therefore, your opening argument, for example, should be equally open to possibility (e.g. don't box yourself in by saying, "Yes, my client did it, but he was desperate..." if there's any possibility that he won't be identified, because your summation will be, "Yes, I know I said he did it, but obviously he didn't...")
The theory of your case is going to shape everything else from here on out. In the voir dire for the bank robbery case, I want to ask the potential jurors questions about identity and misidentification. Likewise, in my opening argument, I'm going to say things like "That's not my client on the video," not "The bank wasn't robbed."
With your theory of the case figured out, you can almost see your summation forming. And, in fact, some lawyers start by writing their summation. I don't literally write mine out, but it's useful to at least think about. In formulating your summation, you start to see what points you want to make with each witness, which will help you write your questions. In the bank robbery example, I might imagine saying my summation, "Ladies and gentlemen of the jury, the teller told you that she only had one second to see the robber. She told you that he had a mask over his face and she couldn't see his face at all. She couldn't see his hair at all. She couldn't see his skin tone. She wasn't sure how tall he was. She didn't hear his voice..." You start to get a sense of the questions I'll need to ask the teller to elicit this information.
One thing that I've often been asked is whether I write out my questions, or work off notes or an outline. It depends on the witness, but generally I write everything out.
When don't I write anything? When it's going to be quick, simple and painless. Maybe the prosecutor is putting this witness on the stand to get one fact into the record, and it's a fact that I don't dispute. So, maybe I won't ask anything. Keeping my theory of the case in mind, maybe I want to get one additional fact out of this witness, I can probably get away without writing anything out.
Sometimes I just don't know what the witness is going to say, and therefore, I don't have much to write out. It's a skill to be able to think on your feet, and figure out what you want to ask on cross-examination before you start to look silly standing there, stalling.
Why do I write out my questions the rest of the time?
First, it helps me to remember to get out all of the information I need from each witness.
Second, it reminds me to use leading questions on cross-examination and to ask good open questions on direct examination.
Third, to set a witness up for impeachment, based on a prior statement, it is best to use the language from their original statement. I write out my question, using the language that the witness used in their prior statement, and then I make a note to myself where to find their prior statement (e.g. "You only saw the person demanding the money for one second. (hearing transcript, page 30, line 5)")
Going back to the binder and redweld camps, I make sure I have my questions and arguments printed and filed in each folder or divider, along with any impeachment material, exhibits I want to show the witness, or items I might use to reflect his recollection.
I hope this was helpful, I've tried to be vague without being too vague, give examples without being too elementary. If you have any questions or ideas, I'd love to hear them. Coming up in Step Four: Thinking About Logistics.