How do prosecutors prepare to cross-examine defendants?

Question

Prosecutor Peter is preparing for a trial. Defendant Dexter may or may not take the witness stand. Peter has no idea what he would say if he does (no disclosure obligation on the defendant) but still needs to be ready.

So, Peter thinks up a few lines of defence that Dexter could possibly adopt on his evidence-in-chief and prepares separate sets of cross-examination questions for each.

To each question Peter expects one or more possible answers. Every subsequent question will often depend on what the answer to the previous one was. Peter aims to plan the questioning such that inconsistencies in Dexter's answers surface as soon as possible so that it becomes apparent that he is lying and Peter could just say "no further questions, Your Honour".

Is that roughly how prosecutors prepare? Do they draw flowcharts? Do they use some special methodologies? Or do they just fly by the seat of their pants?

Answer 1

I'd imagine that testimony from the defendant is rare enough that in the majority of cases, prosecutors do not meaningfully prepare for a cross examination.

To the extent they do, I'd expect the preparation is similar to that for basically any other witness. So I wouldn't expect complicated flowcharts, because the general rule at trial is that you only ask questions whose answers are both known and helpful.

So if I need to place the defendant at the OK Corral at 3 p.m., I'm only going to ask him where he was at 3 p.m. if I have evidence showing that fact is true -- maybe he gave a written statement to the sheriff, maybe he posed for a daguerrotype, whatever. I expect him to deny it, so I don't ask the question unless I have evidence more convincing than his denial.

In this way, a defendant -- like any hostile witness -- is used less to provide any facts of their own, but rather as an involuntary narrator of my own story, authenticating evidence and validating the facts consistent with my theory of the case.

Answer 2

I'm not familiar with criminal court procedure, but in a civil case the most common approach I've seen and used is:

• Go through both sides' statements of case (whether or not that includes a defence) and pick out any points which you think the witness may be able to usefully speak about.
• Make a brief list of bullet points of things you would like to ask. Alternatively, write them directly on the statement of case so that you have a narrative flow for each witness.
• Take your time and feel free to pause for thought between questions. Judges usually prefer you to take things slowly anyway so that they have time to make notes.
• Know the case inside out so that you are able to react on your feet if the answers take you in an unexpected direction.

The last point is the most important in my opinion. If you know your case thoroughly you can react to pretty much anything. If you're overworked or lazy then you are more likely to be caught by surprise and make a mistake.

Sometimes you can get lucky and you'll realise the witness is leading themself into a trap which you can take advantage of by asking a series of questions, the final answer to which weakens the opponent's case but is hard to avoid due to earlier answers. It's hard to plan for these, but sometimes you might devise ways to achieve this in advance.

Answer 3

The answer is most likely "learn some basic interrogation techniques, have a good memory, and be able to think on your feet." Of course the prosecutor will have pre-loaded his/her own memory with the written evidence available to him/her.

The object of cross-examination is not so much to obtain facts as to lower the witness's credibility. Police officers are trained to "cross examine" the people they interview "cold" with no preparation at all. The basic technique is to repeat questions in different ways until the witness contradicts themselves, and then "go for the kill."

If a defendant knows they are guilty but is trying to convince the court they are not, most likely they will have constructed a story about a false version of events and attempt to stick to that story. The strategy fails because the story only covers a limited version of "the whole truth."

For example in an interview or cross examination about a motor accident, suppose there is a detailed (but fabricated) story about how a blue car was driven. A common technique would be to move on to a different topic for a while, pause for thought, and then ask "Remind me what happened when the red car did such-and-such." If the interviewee's fabricated version of events only includes the existence of one car, they will probably ignore the word "red" and assume the interviewer said "blue". When the follow-up question "so are you now saying there were two cars involved, a red one and a blue one?" is asked, the fabricated story will quickly fall apart, one way or another.

(Full disclosure: I am neither a police officer nor a barrister, but I have been trained as an auditor, and I have served on juries. The use of the same questioning techniques was rather obvious!)