Polarizing Question

Two of the most difficult situations a facilitator faces are false agreement and false disagreement (ViolentAgreement?). This is a technique to use when you see false agreement.

When I saw two people who thought they agreed, but I thought they didn't, I always used to point it out by restating their positions in such a way that the distinction became obvious. "Now, John, you just said ... but Bill here, you said ... These seem to be opposites."

This technique doesn't work well at all. They usually reply with 'That isn't what I said at all!'.

Now what I do is find a YesNoQuestion that I can pose to both people that they will answer differently. There is always a pause after the two answers. Then the discussion can resume productively. Either that or they answer the question the same way and I learn something.

I had a couple of good examples recently. One was when a not-for-profit asked me to give them a deep discount. I asked, "Would I just hand you that much money in cash?" Another was when favoritism was being shown someone at a dance studio. I asked, "Would someone from a different family be treated the same way?" In both case, posing the question clarified the situation.

-- KentBeck
Could it be that sometimes you need to question purpose. There could be disagreement about How, but there can be a different why too. Do they think they agree, or do they pretend they agree. Is agreement or disagreement about means, about solution or about purpose? Sometimes we need a series of subsequent Why's as question I think. Remind me to elaborate the InputOutputManagement? pattern (know uses Vatican, Sabena, Argo, European Commission). You just made me think of this. -- MartineDevos
Does a question function better than a statement when one is attempting to communicate some point? Mortimer Adler uses the term "maieutic questioning." The word Maieutic comes from the Greek word maieutikos which relates to midwifery. Webster's define's the word as "relating to or resembling the Socratic method of eliciting new ideas from another." --JoshuaKerievsky
A thought on why Kent's use of the technique might not work well: There may be overt or covert reasons for a false agreement. Management may have issued a "thou shalt agree" edict, requiring that people at least act like they're in agreement, or people who have bonded through some shared experience (e.g., a tough project) may have papered over their disagreements, covertly agreeing to not disagree.

In both cases, a "John said A, but Bill said B" approach puts the focus on the people doing the not disagreeing. If there is a false agreement, this approach--especially if done in public--may force defensive reactions. Shifting the focus to either the thing or the questioner might help. "Help me understand this. I'm not sure if it is this way or that way. Am I confused?" often works for me.

Asking questions almost always works better than making assertions. The trick with Socrates and Kent's questions is that the person asking the question has some idea as to what question to ask next. Socrates would ask a series of proximate questions leading up to what he wanted to explain and then say, "Now note, I did not answer any of these questions, you thought up this idea all by yourself!"

Asking questions gives you, the questioner, the option of being wrong without actually putting anyone on the spot. From the answer, you either confirm what you knew, learn what the next question might be, where the answerer is coming from, or learn that you were off base. With an assertion, you don't get that kind of coverage.
 -- AlistairCockburn

I have the opposite problem. I say, "Why did you do X", because I can think of half a dozen reasons and I want to know which ones are most important. People interpret me as saying I think X is a bad thing to do, that I don't know any reasons to do it. So yes, questions are leading, and that can get in the way. People won't always take them at face value. -- DaveHarris
I largely agree with DaveSmith's reason that Kent's previous practice did not work well. After something is (allegedly) explained, a comment "... you said ... he said ... these seem to be opposites..." sounds like a challenge - "You guys really don't agree, although you think you do" or "You guys don't really understand what you just said." As Dave says, this leads to a defensive reaction as the parties try to prove that their prior belief (that they agreed) is true.

The YesNoQuestion does not challenge. It simply asks for more information, making no assumptions on what that answer will be. It also very concisely demonstrates that their prior assumption of agreement is wrong - *without claiming it to be* As a result, they do not feel a need to defend themselves from the facilitator (which generally indicates a facilitation error).

The issue raised by DaveHarris is very similar in that the question, "Why did you do this?" is a challenge. Some alternative approaches might be to ask what would happen in a particular circumstance (if you know of a problem), or even simply to ask, "Does anyone see any problem with this approach?" (falling back to the first alternative if no one does).

Yet another possibility might be to ask why a particular approach was *not* followed - but only if there is some reason for all concerned to have expected it to be followed, such as - we just bought this tool, why aren't we using it? or, didn't we agree to use this other technique? What are the conditions under which we should use the approach you did, instead?

-- RussellGold
I've had the same experience that Dave and Russell report: it seems that asking "why" is very challenging. Particularly when you use that tone of voice that implies "you idiot". Almost any alternative formulation would be better.

I tend to ask about the artifact a lot. When you say "please tell me about this thing here.." people will tell you a lot about their intentions without being asked directly. Sometimes they tell you too much, or things that you already know. One way of out this is to LetTheLearnerGuideTheFlow. This doesn't have much to do with facilitation as much as learning rapidly. -- MichaelFeathers

Concerning the 'why did you do X' problem DaveHarris points out. When I am really trying to learn which of the several alternative explanations people may have had, and when I do, in general, think that their actions or choices were justified, when I just want to learn their particular justifications, I'll say something like an enthusiastic -

"X was brilliant!!! How did you come up with that idea?"

It's a little over the top, but I've rarely had anyone complain, "no, no, it wasn't brilliant at all." And I almost always get the details of their thought processes, because now they have a reason to be proud of them.

Try it. It is kinda fun. (if it is impossible to say that x was brilliant, find someway to preface the question with the information that you approve of x)

-- TonyReno
There is another thing that works in some cases: Just ask the proponents about the strengths of their solutions. Then they can talk about the advantages (which gives you the opportunity to look for weaknesses). Once all advantages have been discussed, take the list of potential weaknesses and formulate questions in order to find out how they are addressed. -- ManfredLange
I recently did a practitioner course in NLP (NeuroLinguisticProgramming), which has a lot of question-based techniques. However NLP almost never uses "why" questions, since they almost always result in a defensive reaction. Instead of asking "Why did you do X?" asking something like "What did you want to achieve by doing X?" will get a much more useful response, even though the two questions mean almost exactly the same thing.

It struck me that there are a lot of similarities between NLP & XP - in particular the emphasis on feedback and metaphor.

-- DaveKirby

I'm not completely convinced that asking someone "Why did you do ..." and "What did you want to achieve ..." is the same, i.e. it wouldn't give the same information. Why - questions the strategy, while What - questions the motive.

-- NiclasOlofsson

Niclas, you're dropping context. It's not "why" versus "what", it's "why did you do" versus "what did you want to achieve by". Those extra words are important, and to this native English speaker, they really, really do mean the same thing. --Samuel A. Falvo II

On questions: I think it's sufficient to ask about a decision by finding out if you understand the decision. One that I'd like to ask the ATL folks: "When you unioned this pointer to T with a pointer to a pointer to T, does that make the memory management faster in the typical case?" I'm not sure why they did it, but if I come up with a question that has a yes/no answer that also begs explanation, it's not going to make someone defensive, and it may encourage teaching. (Another question: Why are Microsoft's programmers pathologically creative? Look at the definition of CComUnkArray<> and CComDynamicUnkArray in <atlcom.h>.)

Another comment up there, about preceding q/a sessions with "brilliant idea of yours," made me worry in a way. I don't think that a person who doesn't do this innately should try to learn it, because it makes others think that that person has been reading management textbooks. Also, what if your goal is to challenge the idea and perhaps change the code if not satisfied? Then you get into the PointyHairedBoss territory of contradictory management. -- JohnDuncan


View edit of August 16, 2010 or FindPage with title or text search