In 1980, Chris Argyris published a short article called ‘Making the Undiscussable and its Undiscussability Discussable’. Argyris wrote broadly and for decades about organizational behavior and its consequences, but this is the article with the catchy title and the fun terminology so it’s the one people remember. Let’s start there.

(Access the article via JSTOR or archive.org.)

Undiscussables 1980

Argyris wrote that as individuals enter organizations they bring

a set of values, action strategies, and skills that lead them to respond automatically to threatening issues by ‘easing in,’ ‘appropriately covering,’ or by ‘being civilized.’

Any way you wish to describe the actions, they add up to making threatening issues undiscussable and then to making their undiscussability undiscussable. The organization may not be the culprit; it may be the victim of the individuals who work within it. However, once the victim, the organization may collude to maintain and reinforce the problem.

The article uses undiscussability to diagnose mismatches between various learning strategies available—what Argyris called single-loop vs. double-loop learning—and their (mis)application to various problems that arise. This is valuable and good, but it feels incomplete.

Undiscussables 2025

At 45 years’ remove from the paper, there has been a change:

• Back in 1980, Argyris approached undiscussables as emergent, unwanted behaviors that a wise organization could route around in order to achieve shared understanding and clarity of purpose.
• In 2025, individuals bring their own perspectives on undiscussability into an organization, per Argyris. But the typical organization has also established its own set(s) of undiscussables. To what end?—as a means of control.

This is not solely a result of the present political circumstance. Although authoritarianism trickles down in a way that most things do not, what we’ve seen in the last year in the workplace is a change in degree, not in kind.

The widening circle of undiscussability has consequences.

1. The further up someone is in the org chart, the more fabricated (as in: filtered, or ‘eased in’ to acceptability) the information they receive. This huge disconnect helps nobody, as Argyris discussed in the 1980 paper.
2. Conversely, lower on the org chart, a simply outrageously tremendous amount of labor goes into producing these limited, acceptable, discussable messages.
3. Things go worse than they otherwise might. Opportunities and ideas are constrained. Failures take longer and cost more. People are excluded.
4. And you aren’t meant to talk about any of it. It’s all undiscussable.

Courage & kindness

I bring all this up because I want you to have courage. I want you to know that if you’re feeling this pressure, you’re not alone. I don’t suggest you stand up and shout out the organization’s undiscussables. But maybe write them down somewhere, on a 100% offline piece of paper, and think about how their mere, invisible, undiscussed presence benefits the organization.

Returning at last to Argyris’ formulation of undiscussables as ‘threatening issues’: what are they afraid of?—what are they threatening you with?—and what kindness or stout-heartedness might you extend today, tomorrow, and again the day after, to others feeling the same sense of enclosure?

I’m starting a new client project next week. Now, I’ve been at this for a long time—it’s consulting project #74 (& yours could be #75!). Along the way, I cobbled together a short checklist for the first days of a new engagement.

Here it is:

1. understand the project,
2. understand the client, &
3. understand the work.

Sounds simple—and it sometimes is—but I’ve watched skilled people create problems for themselves by rushing in. Hell, I’ve been one of those people.

Understanding the project

Take as given some sort of plan or statement of work or agreed-upon outcome—which is great, but not what I’m interested in up front.

Instead, the very first thing I wonder is: who wants the results of this work to be good (also, what might ‘good’ mean for them)? And the second thing is: who is worried it will be bad (and what are they worried about)? I might host a premortem in order to get people talking specifically about the bad things that might happen. This helps me suss out who cares about what, and who else should have been included that hasn’t been (yet).

I am also drawing pictures in order to answer questions like: why this work? and why now? A round of ecocycle planning can bring this to light. Another approach is to map the value chain, which will reveal some dependencies, possibly hidden, and clarify what’s super important and what’s not.

Understanding the client

How does a client want to interact with the consultant or the consulting team? ← You would be amazed at how often consultants do not ask this.

I follow Jerry Weinberg’s advice (“Your methods of working are always open for display and discussion with your clients”) but it’s an invitation, not a demand. Some clients want to work “at the elbow” and participate in the mess from beginning to end. Others want to check in weekly, or monthly. Dial this in and you’ll know whether to work with the metaphorical garage door up or down.

When getting to know a new client, I will pay attention to what they react to in materials (documents, plans, etc.) and what it’ll take to get good feedback on those materials. I remember the client for whom we did a full style adherence pass on even the roughest drafts and sketchiest ideas before putting thing in front of them—because otherwise we’d get a sequence of comments about images being imperfectly aligned and nothing else. And that’s OK—giving thoughtful feedback is hard and people are busy. Sometimes clients, or their teams, need an idiosyncratic feedback container & review process (example) in order to elicit the right kind of notes at the right time.

At the same time, I figure out how to make myself easy to fire. I recently ended a project 2 months early due to my client’s budget getting jerked around. This came at precisely the worst time: I was close to wrapping up, but not quite finished. Even so, I was able to hand things over in an orderly way. I’d kept things organized as we went, so the various transitions (or “knowledge transfers,” to use the unlovely jargon) were easy to make. I could focus on communication and relationship and problem-solving, rather than scrambling to claw half-baked work items together.

An old consulting boss said she never wanted her clients to create a dependency on our services. There’s always more work to do, and people mostly remember how they felt at the end of something, no matter how miserable the middle portion was. Being easy to fire means that I can always find time at the end of a project—even when they end early or suddenly—to make the ending nice and calm and appropriately bittersweet for everybody.

Understanding the work

Finally, I will begin to indulge my curiosity. Not just in the work specifically entailed in the project, but in everything surrounding it. This is the fun part! It’s also where, as my friend Andre says, “every project is a lean project.” I start this by going to where the work happens, and spending time with the lowest-status people doing the work, and paying attention. The point is to learn the process, notice where it breaks down, and find the places where success hinges on tacit knowledge.

↑ Read my field guide on the 6 MISSED wastes for more on how to do this.

When people tell me things like “we don’t have a process“ or “every x, y, or z is different,” I’ll break out a work-unit routing analysis (WURA, sometimes also known as product-quantity-routing or PQR analysis) worksheet and become a student of the variation I see and hear about. This gets me on the way towards sorting things into value streams or some other aggregation.

Despite the experience or expertise that the client presumably hired me for, in these moments the value I bring is that I can look with a beginner’s eyes and wonder with a curious mind.

Last but not least: doing the project

But that is—as they say—another story, and shall be told another time.

People write for lots of reasons: To become famous. To make money. To sell. To connect. To join. To have written. To create social proof. To influence. To communicate. To make a world. To right a wrong. To increase the n. Because everybody else is doing it. Because someone else said they should do it.

Today, 23 reasons I think you should consider writing online. Pick one reason, pick them all, make them your own.

Reasons

1. To allow a small number of as-yet-unknown, possibly very consequential people to feel as though they’ve gotten to know you a bit, before you ever speak. (I’m a consultant. 3 out of 4 interviews I have with prospective clients involve them bringing up something they read or listened to from this web site. I enjoy that.)
2. To make your first conversations with certain new friends immediately engaging and specific, since they will have already apprehended your bullshit and felt there was something there.
3. To disqualify people who don’t grok what you are up to. They will glance off the edge of your atmosphere and float somewhere else.
4. To sort things out, to somehow make sense of an absurd world, and cram the results into enough of a structure that it becomes hopefully intelligible to at least one other person.
5. To dispel the curse of knowledge. You know much more about certain things than most people, and your life experiences mean those things have become connected—intertwingled?—in contingent, weird, productive ways.
6. To dispel the curse of knowledge. You probably don’t realize how much you know about the stuff you know about. Lay it all out on a page and, goddamnit, turns out you know a lot. That in itself is good to know.
7. To dispel the curse of knowledge. Here is a super powerful capability: to be able to borrow the perspective of someone who doesn’t know the same things as you do, and support them as they develop their own expertise.
8. To get clear about the language you want to use. And I don’t mean cussing (you must swear at least a bit in your writing so people will know they are dealing with a human being and not some tediously apologetic autocomplete LLM). I mean knowing what things you are thinking about, and what you would like to call each of them.
9. To identify your lil’ shortcuts, jargons, and irritants.
   (For me, these include the words “just” and “here.”)
10. To find out where you are wrong.
11. To find out where you are right.
12. To eventually give each idea its own URL, which is cool.
13. To take these ideas—now having their own URLs—and e-mail them to those who want to read by email, publish in an RSS feed for those who want to read by feeds, and the like.
14. To maybe even circulate these URLs across the various algorithmic hellscapes (LinkedIn, Twitter, etc.), giving people something to glance at between the advertisements. Your items are errant kibble in the feed, but you can automate the postings and just—see #9—let the machine operate.
15. To locate the people who are your people. Bring them in, wherever they are. One person a month is fine. This is the real deal.
16. To have a very slow conversation with a friend over the span of months or years.
17. To accumulate a “swipe file” you can borrow from when called upon to present, speak, convince, design, etc.
18. To draft a book, one page at a time, with the garage door up.
19. To create the conditions for eventually revisiting your old stuff and noticing with horror how your thinking has changed over time.
20. To get other people to tell you what to read next: this link, this article, this book. You’ll always have a backlog.
21. To learn.
22. To practice.
23. To figure out who you are, so you can become that person.

I’m a digital pack rat. So is my friend Lee. We studied human-computer interaction in grad school back in the day; the entire program was a warm, nutrient-rich incubator for nourishing little digital pack rats. Well, we were talking and uncovered a new rule.

I give you Lee’s Law:

“Every idea should have a URL, bonus points if one does not have to make the URL oneself.”

Here are three sites that satisfy Lee’s Law. Three different bodies of work from three distinct perspectives. What makes them awesome is this: they provide just enough information to make each idea useful. Find the page you need and use it yourself. Or share it with somebody. Try presenting directly from the page.

Bookmark these and come back to them at least once, and—congratulations—you're a digital pack rat too. One of us!

• Liberating Structures: Facilitation approach and methods. Indispensable. If I could only keep 5 web sites this would be one of them.
• 18F Methods: Methods gallery for human-centered product or experience design. I've run entire projects straight out of this playbook.
• Laws of UX: For those times when you remember what the thing is, but not what it's called.

Find your main point: one thing you want people to remember. Begin the presentation by making your main point as quickly and plainly as you can.

Then, state the main point again. Elaborate on it this time, with a little more context. Give the two most relevant examples you can.

Next, revisit the main point. Explain how it is applicable or meaningful to your audience, based on the elaboration you just did and your understanding of their situation. This is where you can get ultra specific and pile on the details. Nerd out. This works because you’ve built little gathering places into your presentation. When people get distracted or you do a shitty job explaining something, some of your audience’s attention will wander. That’s OK. Each reiteration of your main point is an opportunity for people to get back on board. It is a starting point everyone will be familiar with—since it’s your main point.

Finally, revisit the main point. Remind everyone what the main point is, and what it means for them.

Conclude with a quote or a story that emphasizes the main point.

Having a hard time organizing a talk? Plan it backwards.

1. Think of how it will end: the capacities, agreement, and energy your audience will have. That’s your ending.
2. Strip away some detail, and slot that before the ending.
3. Strip all but the main idea, and that’s your opening.

First way: try out guesses with small, reversible tests

Try out changes on a small scale—involving a few people for a short period of time—and in a way that’s easy to bail on.

Things tried out this way tend to be crappy or prototype-y. Good:

• Use a piece of paper instead of a fancy app.
• Make a spreadsheet instead of a report.
• Draw a table and figures on the wall instead of a spreadsheet.
• Do it manually today. Automate later.

Small: you need only enough work to figure out if your guess was any good. Don’t dump resources or spend time rushing towards dead ends.

Reversible: work this way to take the sting out of failure. Idea didn’t work? Stop and try the next one.

Test: once you can point to an initial success, gather resources and proceed with something more elaborate.

Second way: a logbook of guesses

Once you start testing your guesses, you’ll find that you make a lot of guesses, and run a lot of little tests. So keep a log. Write down:

• Your guess,
• What you thought might happen, and
• What actually happened.
• Was your test bigger or smaller than it needed to be?

Practice this and your guesses will improve over time.

Don’t take my word for it—try it in a small, reversible way, and see.