Dave Dolkas
Dave Dolkas


The Art And Science Of Investigative Questions: Part 5 (Level One and Two Questions)

In this five-part article series, we have discussed a number of ideas about how to ask investigative questions when lawyers must elicit information from a source or a witness. This final article addresses Level One and Level Two investigative questioning.

Level One questions are planned before the interview or examination starts. They are questions you'll likely ask the source or witness. The more difficult the issue or topic, or the more sophisticated the witness, the more preparation is needed. Begin your questioning with areas of agreement because accountability ultimately comes out of agreement, not fireworks. Agreement means asking questions that cause the source to establish foundational facts through his/her answer. Agreement doesn't mean just reading out research or statistics. It means using the question to make the source take ownership of facts.

Level Two questions exploit the answers given by the source to Level One questions. The Level Two question is asked spontaneously based on the source's answer to the Level One question. Leveraging the source's answer to a question is akin to a judo-like approach. (The Japanese word judo means: the soft or gentle way or art.) For instance, if the source responds to a Level One question with an answer like: "There were some circumstances we had not anticipated," then the Level Two question is: "What were the circumstances you had not anticipated?"

The point is that the examiner uses the power of the answer, rather than the power of the question. In answers to questions lies the power an examiner needs to form other questions and extract critical information.

Let's return to Mike Wallace's interview of Nancy Reagan (discussed in Part 3). Wallace posed this hanging question: "You're going to be in Japan and I'm told it's a $2 million two weeks." That's a pre-planned Level One question intended to extract an agreement. Reagan answered: "There's going to be two of us and they're working us like crazy."

Some follow-up Level Two questions could be:

Q. How are they going to work you and your husband like crazy?

Q. How do you respond to criticism from our viewers that you're profiting from your former positions?

Q. Why isn't that a fair criticism?

Whatever the Level Two question is, it's framed in response to the answer given to the Level One question. You don't know what that Level Two question is until you hear the source's answer to the Level One question. You can map out your game plan before the interview begins by anticipating the source's answers to your Level One questions. But you won't select the Level Two question until you hear the answer to the Level One question.

Returning to the Wallace interview of Roger Clemens (discussed in Part 1), here again are the questions Wallace asked Clemens and Clemens' answers:

[Wallace] Did your former trainer, Brian McNamee, ever inject you with anything?

[Clemens] Yes, he did.

[Wallace] What?

[Clemens] Lidocaine and B12.

[Wallace] Never, never a human-growth hormone?

[Clemens] Never.

[Wallace] Never testosterone?

[Clemens] Never. Never.

[Wallace] Swear?

[Clemens] Swear.

In an article by John Sawatsky — award-winning journalist and ESPN's senior director of talent development — published a few days before the "60 Minutes" broadcast aired, Sawatsky suggested that Wallace ask Clemens these Level One questions to extract agreement from Clemens:

Q. How widespread has the use of steroids been in Major League Baseball in the past few decades?

Q. How much of competitive advantage do players get from the use of steroids?

Q. How would you describe what happened with home-run totals in the 90s?

Q. What did you observe at the time from being inside the game?

Q. Why did you speak out at that time?

Q. Why would Brian McNamee lie about you when telling the truth about Andy Pettitte?

Sawatsky explained: "These are solid questions that move Clemens into a different space. However, good questions alone do not necessarily pull the truth out of anyone. Successful interviews get people to go further than they planned to go, and rarely come from a planned list of questions, even when the questions are good ones." Sawatsky explained further that interviewing "is a dynamic process involving two basic stages. Stage 1 is planned; Stage 2 explains the moment that Stage 1 produces whenever and however it occurs." (Sawatsky's Stage 1 and Stage 2 are what I'm referring to as Level One and Level Two questions.)

In the interview, Wallace asked the "why" questions (bolded above) suggested by Sawatsky:

[Wallace] Why would Brian McNamee tell the truth about Andy Pettitte and lie about you?

[Clemens] Andy's case is totally, is totally separate. I was shocked to learn about Andy's situation. Had no idea.

Unfortunately, we don't know whether Wallace asked any follow-up Level Two questions because the broadcast cut to a track-over. Using the Sawatsky method, I would have asked Clemens this follow-up Level Two question:

Q. How do you explain to our viewers that you were shocked to learn about Andy's situation when you were good friends, on the same team, and had the same trainer?

That's a Level Two question exploiting Clemens' previous answer and incorporating Clemens' own words into the question. Clemens likely has two choices in how best to answer that question. First, and realizing he may have put himself out on a limb, he could plead ignorance: "I have no explanation." Second, Clemens could bunker in further and try to provide an explanation-for instance: "Andy's injury was a lot worse than anything I ever had."

For the Level Two question in that instance, I'd probably leave the first answer alone and not follow up. If Clemens had no explanation, that in itself would be a revealing answer. If Clemens gave the second answer, then my follow-up would be:

Q. How do you know that Andy's injury was worse than anything you ever had?

The "how do you know that?" question works whenever a witness volunteers an answer that appears to be pulled out of an ear, based on sheer speculation. When confronted with a tough (not tough-sounding) question, witnesses often give answers like this one. The witness' red brain makes stuff up to get out of a tight jam — usually some speculative answer to get into a better light or to try to deflect further tough questions. If you receive an answer like this one, follow up and probe the basis of the purported knowledge using some or all of the five Ws and One H:

  • How do you know that?
  • Who told you that?
  • What do you know about how taking steroids speeds up recovery from an injury?
  • When did you see, read or hear that?
  • Where did you see, read or hear that?
  • Why do you believe that?

Don't stop asking these questions until you've completely elicited the source of the knowledge or established that it's obvious the answer is complete speculation. Doing this will also send the message to sources that whenever they try this tactic, you'll follow up with questions testing every aspect of their supposed knowledge.


Because I've covered a lot of ground in this five-part article series, here's a summary of 10 key rules for investigative questioning:

  1. Think input, not output. You can't blow and suck information simultaneously. And remember, it's not a conversation.
  2. Use the five Ws and one H.
  3. Use the power of the "why."
  4. With your questions, force the source to choose shades of gray, not black or white.
  5. Ask lean, clean and open-ended questions. Avoid asking long, overloaded, close-ended questions.
  6. Probe tough issues. Avoid tough-sounding questions.
  7. Ask short, but probing questions. Don't ask double-barreled questions.
  8. Ask a question (Question = topic + inquiry). Don't make statements containing your opinion or trigger words.
  9. Pre-plan your Level One questions, using the five Ws and one H. Consider the possible Level Two questions to ask by anticipating the source's answers to the Level One questions.
  10. Ask: How do you know that? and other follow-up questions whenever it appears that an answer is made up.

David Dolkas was a litigator and trial lawyer for 34 years most recently with McDermott Will & Emery LLP. He now trains and coaches lawyers on how to MAP (manage, analyze and present) complex cases and how to map careers.

David Dolkas