Being a Devil’s Advocate
In the Monday meeting, one topic that came up from a few different people was the role of “Devil’s Advocate” in a group discussion setting, and the benefits and challenges of that role. I used to take that role a lot, but have recently found value in moving away from it — or, perhaps, evolving it into something new, that I don’t have a name for, that maximizes its value both to myself and my team.
Since I unfortunately missed Monday’s meeting, I wasn’t able to share my thoughts at the time. However, this meant I had plenty of time to mull over my feelings on the issue, and I decided to take the time to share my thoughts more deliberately here, in an attempt to make up for my lack of input earlier.
As I said, I used to very easily and naturally take the role of Devil’s Advocate, — or even more broadly, the role of “Skeptic.” I could always find a compelling counter-argument to any argument presented by others. Believing there is value to considering all sides of an issue, I would even happily argue against positions that I agreed with! I didn’t always pay the most attention to how I presented my disagreements, and as such, this wasn’t always the most effective method of making friends, not to mention getting my positions taken seriously.
Often, I began to notice, one person would say, “I think this is a good idea because of reasons A, B, and C.” And I or someone else would respond, “Well, I think it’s a bad idea because of reasons D, E, and F.” Neither of us was actually addressing the arguments of the other, we were just talking past each other, and it only led to frustration on both sides.
A few years ago, I had a revelation. I was trying to decide which programming language to learn next — Ruby or Python, both of which were rapidly growing at the time, and both of which showed enormous potential. I soon found an article on Python called something like “10 things I hate about Python.” What caught my attention was that it was written by a huge fan of Python, who used it every day and prefered it over any other language he had used.
The articles I was able to find promoting Ruby all had a too-enthusiastic, almost cult-like feeling, written by people wearing rose-colored glasses who were blind to any of the real weaknesses of the language. They could talk only about Ruby’s strengths. Whereas, the experience of seeing someone grounded enough to raise honest and aggressive criticism against his language of choice, while showing implicitly why he felt it was worth using anyway, convinced me in an instant that I’d rather join this rational skeptic on the Python side of things.
This experience taught me a lot about the subtleties of the role of Devil’s Advocate. There’s an idea in the scientific community that the only real way to develop evidence in favor of a theory is to form specific hypotheses based on the theory, and then design experiments that attempt to falsify them. our natural tendency is to only seek information that supports our theory, but what we should really be doing is exactly the opposite — every time we fail in a good-faith attempt to disprove our own theory, that’s when it really gains value.
From this, I learned the greatest value of devil’s advocacy: If you can’t argue convincingly against your own opinion, you haven’t thought through your position thoroughly enough.
Conversely, if you can’t argue convincingly in favor of your opponent’s position, you probably don’t understand why they feel the way they do well enough to change their mind, no matter how strong your counter-argument sounds to your own ears.
This process of arguing in favor of your opponent’s case is called “Steelmanning.” If you can present an argument you disagree with in its strongest form, even better than it was presented by your opponent, you can find in your own mind the mental state in which that argument makes sense. When you then argue against the point from this mental state, it reveals: “I see exactly where you’re coming from, and given that, here’s my concern.” Your counter-argument can be presented in exactly the form that makes it most digestible to the person you’re talking to.
With this in mind, I now try to agree by default with every argument presented. I try to build on the arguments and make them even stronger. Sometimes, I manage to convince myself — I realize I was holding on to an assumption or bias that was holding me back in my own thinking. When I fail to convince myself, I know that I’m now in a strong position to constructively present a counter-argument — or, even more subtly, to present the underlying basis for my concerns, allowing others to follow my train of thought themselves and discover the counter-argument independently!
This isn’t about “walking on eggshells,” or avoiding bruising sensitive egos by couching criticism with complements. It’s about debating in good faith, and always acknowledging that, no matter what my gut instinct tells me, it’s disrespectful to ask someone else to change their point of view without first making a genuine attempt to do so myself.
At that point, even if I do convince myself that my opponent was right all along, I can now return to the technique in the Python essay and critique the position I now believe in. And this is the strongest form of the Devil’s Advocate — if you can argue genuinely against your own belief, and fail to change your mind, then — like the failed attempt to disprove your own scientific theory — you can be confident that you’re on the right track.
So now, when I engage in strategy discussions, I’m more like a facilitator than anything else. I strive to make sure I have heard and understood every viewpoint, and to argue in favor of those viewpoints myself, before engaging in any sort of conflict. Doing so not only results in less conflict overall, it results in much more productive conflict that helps everyone understand everyone else’s point of view and reach a consensus with minimal tension.
