talking sense — teffer.dev

Everyone makes sense to themselves. In our minds, ideas live unconstrained — our own context, our own logic. The moment we translate them into words they leave us. Other people interpret what we said through their own context, their own experience. The ideas land somewhere different from where they started.

In a study at the University of Chicago, psychologists Boaz Keysar and Anne S. Henly paired speakers with listeners and found that 46% of the time, when the speaker thought they had been understood, the listener had completely missed it. Speakers’ confidence in being understood had zero correlation with whether they actually were. Cognitive scientists call this the “illusion of transparency” — we overestimate how obvious our meaning is to others. Combined with “egocentric anchoring” (our brains anchor to our own context and feelings), it means we are systematically wrong about whether we are getting through.

I have sometimes been described as someone who is an empath or exhibits high EQ. I owe this to my executive coach. My coach taught me a leadership / communication system called “sense-making”. Sense-making has 3 main parts: 1/ structuring conversations, 2/ giving good information, 3/ listening for understanding to create the conditions for collaborative conversations and outcomes. These are the concepts from Talk Sense: Communicating to Lead and Learn by Barry Jentz. This post will focus on the last one, listening for understanding.

In one coaching session that still stands out for me, my coach guided me to listen out loud using phrases like “I am hearing you say X” — but don’t just mirror what they are saying. You haven’t successfully made sense until you can repeat the other person’s sense-making system out loud with a meaningful insight attached. The most powerful question is “have I got that right? Am I missing anything?” This allows you to validate your sense-making.

tough, nice, or sense

Jentz argues that most leaders, when communication isn’t working, oscillate between two failure modes:

Talk tough. Get more direct. Push harder. Re-state expectations more firmly. “Let me be clear about what I need from this team.”
Talk nice. Soften the edges. Lead with empathy. Validate, then nudge. “I hear you, and I think we should also consider…”

Both feel like leadership in the moment. Neither tends to actually move performance, because both are still operating inside the leader’s own sense-making. The leader hasn’t done the harder work of understanding what’s making sense to the other person — they’ve just changed the volume on their own message.

Jentz’s third mode is talk sense. Stop adjusting tone and start adjusting attention. Get genuinely curious about the sense the other person is making. Surface their reasoning before you respond to it. Build shared understanding before you push for action.

This is, in my experience, the harder thing. It is also the only one of the three that consistently works.

reflective vs. reflexive

The mechanic underneath “talk sense” is what Jentz calls reflective communication, in contrast to reflexive communication.

Reflexive is what most of us do most of the time. Someone says something. We react inside our own model of the world. The reaction comes out as our next sentence. The conversation is two people taking turns talking past each other in their own private logic.

Reflective is slower. Someone says something. You pause and try to construct their model — what assumptions are they making, what data are they seeing, what concerns are driving this — before you respond. You sometimes name what you think they’re saying and ask if you have it right. You make the gap between their sense and yours visible, which is the only way to actually close it.

Reflective communication is uncomfortable at first because it feels slow. But it gets faster with practice. It’s also the only kind of conversation where the people on the other side actually update their thinking instead of waiting for their turn.

what i try to practice

I have made significant progress applying sense-making to conversations, and I am still working to get better at it. I’m strongest at this when I go into a conversation expecting it. I’m weakest when a difficult or high-stakes conversation arrives uninvited — when I haven’t had a chance to mentally pre-load the stance. I try to create an “airgap” — take a breath and move into sense-making.

Some other tactics that help me make sense:

Assume their sense is coherent. If someone is doing something that seems wrong from where I sit, the question isn’t “why are they being unreasonable” — it’s “what does the situation look like from where they’re standing that makes this the right move?” Almost always there’s a coherent answer, and it usually changes my read of the situation.
Ask, then ask again. When someone gives me a position, I try to follow with “help me understand what’s leading you there” before I respond. The first answer is usually a position. The second answer is usually the actual reasoning. The reasoning is the part that’s interesting.
Name the sense back. “What I think I’m hearing is X. Is that fair?” It feels redundant. It is not. It surfaces gaps in interpretation while they’re still small enough to fix in a sentence.
Watch for the urge to convince. When I notice myself building a counter-argument while someone is still talking, I’m in reflexive mode. The fastest way out is to put the counter-argument down and ask one more genuine question.

None of this is a script. It’s a stance. The stance is: my job in this conversation is to understand the sense you’re making before I push my own.

The best rule my coach taught me to tell when I am not listening is simple: If the other person repeats the same thing three times, you are not listening. The only way to break the cycle is to listen.

the ai overlap

When I prompt an agent and the result is wrong, the question is rarely “is the model getting worse.” The question is “what sense did it make of what I asked, given the context I gave it?” Because the agent, like a person, made some sense of my instructions — and if the output is wrong, my prompt was the part that didn’t make sense the way I assumed it would.

The discipline is the same as with people: get curious about the sense being made before you blame the result. With agents this is faster — you can ask, you can iterate, you can show your work and theirs. With people it’s slower and more important. But the move is identical: assume the other side’s interpretation is coherent, surface it, and then close the gap. Asking follow-up questions, and asking what context I could provide to get better output, have been invaluable habits for me as I work with these tools.

Sense-making, with humans or with agents, is the same move at different speeds.

making sense

My coach helped me become a better listener, a better collaborator, and a better sense maker. He wasn’t telling me to be clearer. He was telling me clarity isn’t a property of what I say. It’s a property of what gets received and reconstructed at the other end.

The work of communicating well is mostly the work of getting curious about that reconstruction. Not adjusting tone. Not picking better words. Curious about the sense the other person is making, and willing to slow down long enough to see it.

One night during the pandemic while I was speaking with my coach, my wife walked in and thanked him. She said that she wasn’t sure what we were working on, but the difference in my ability to listen and understand at home was unmistakable.

This is the most influential leadership concept I have learned in my career. It feels uncomfortable at first, and gets easier — especially when you watch your team start to do it back to you. I hope you give sense-making a shot. It’s worth it.