Every day I see people trying hard to avoid meetings. I hear it from teams (“We’re cancelling this so we can get some work done”). A common theme on social media laments the challenge of balancing meeting and work time. Meetingphobia is deeply embedded in tech’s prevailing mindset: for example, in 18F the 👼 emoji is used for meeting cancellation because, “every time you cancel a meeting, an angel gets its wings.”¹

I get it. Meeting culture sucks. It’s too easy for people to thoughtlessly take each others’ time, occupy standing slots, show off with performative teamwork, and generally suck your energy. Meetings feel like dead time. Meetings are time spent with people yet strangely devoid of social gratification. Meetings typically bore most participants — the greatest sin in knowledge work — and when they’re over, nothing has changed except us all being that much closer to retirement.

Really, I get it.

But what if, hear me out, what if the *only* work that matters in a knowledge economy happens when we are together? What if the reason we can’t seem to fix meetings is that we’re mischaracterizing “the work” in the first place? My eventual goal is to challenge the trope that meetings aren’t work, but first we need to take a tour of some big ideas…

Knowledge work, virtue, and dignity

The term “knowledge work” was first coined by Peter Drucker in his provocative book Landmarks of Tomorrow (1966). He needed a way to pinpoint what was different about 20th century work, what it was that was rendering the prevailing styles of management obsolete. The important work of tomorrow’s (now today’s) organizations would be centered on good decision-making. Volumes of production would no longer be the benchmark; success would be measured by outcomes, no matter whether you were in business, government, military or social spheres. Drucker tried to shift attention from doing things right to doing the right things. Sound familiar? Bear in mind, this was almost 60 years ago.

Although “knowledge work” gets used as a badge to assert status over manual labor, the term usefully indicates where the dignity exists in the work. Manual labor has dignity in the physical: in immediate, tangible change to the world. Knowledge work doesn’t have that, and let’s face it, a lot of knowledge work has only a tenuous connection to external value at all.²

The dignity of knowledge work is rooted in our dignity as knowers, a concept from a branch of philosophy about the ethical implications of knowledge and belief. Success in knowledge isn’t about the facts we know, but by how good we are at judging the truth of uncertain things. Drucker measures this in the outcomes of decisions, but for philosophers, good judgement is a worthwhile aim in itself to get along better as individuals and society.

Developing practices to improve our ways of knowing is by no means a new concept; Aristotle went deep on this , cataloging the “intellectual virtues” and describing reflective work to improve knowledge practices. In a healthy workplace, the whole system promotes higher-quality knowledge production, above and beyond what any individual could achieve alone.

Why do we never seem to learn?

My own journey through these ideas started with the puzzle: why do groups of smart people often act with such apparently low collective intelligence?

Unlike most people in the tech industry, I’ve stayed in all my jobs a long time, between 4 and 10 years. Most of my colleagues have been smart, thoughtful, curious, and kind, and yet it has always felt like we were doing a bad job with learning as an organization. We kept repeating large-scale mistakes, and it was the exception rather than the norm when hard-won experience visibly fed back into strategy or process. I’ve since learned that it’s really, really common to feel “my org just can’t learn.”

One possibility is that the orgs are doing great, it’s just our perception that’s wrong. But judging by the 450 pages of dysfunctions cataloged by Chris Argyris in On Organizational Learning, with such upbeat sections as “How Professionals Avoid Learning” and “Defensive Reasoning and the Doom Loop,” both individuals and organizations do typically resist learning. Individual and collective reflexes for self-preservation kick in to prevent real change.

Moreover, although individual learning (already hard enough) is necessary for organizational learning, organizational learning is not a simple by-product of individual learning. It’s significantly more complex for groups of people develop whole new theories-in-use and behaviors to better carry out the organization’s purpose. The interplay between individual and organizational learning makes it a daunting subject to methodically unpick, but at least you can be satisfied that it’s genuinely hard for everyone, not just you.

Learning is social, truth is tribal³

So what does work for learning? Condensing right down,

• A good individual knowledge worker has information and experience in the domain they’re making decisions about.
• A great knowledge worker has a solid practice of learning, demonstrably improving in effectiveness over time.
• The best knowledge workers have a social reflective practice to acquire more diverse input, do a better job of interpretation, and actively find errors in thinking.

Annie Duke describes this brilliantly in Thinking in Bets, where she shows the value of group reflective practice, in her case proven out by winnings at the poker table. Duke’s gambling experience is a nice context to learn about learning, because it has a simple success definition ($$$) and short feedback loops. Even in a game as apparently simple as poker, top players need to form new interpretations and hypotheses which can give an advantage. A tiny improvement on probabilities pays off. As Duke demonstrates, consistent winners are embedded in groups that work together to improve their abilities to judge the truth of uncertain things.

The challenge of judging isn’t specific to the bluff-and-misdirect world of poker; “true enough to act on” is shaky even in the places where we expect it to be solid. For example, the science we’re taught in school has shared standards for validity; concepts like statistical significance and confidence intervals are complicated, but in a reassuringly mathy way.

As you keep going with science, though, you find that “truth” is very much not a solid concept. The ongoing (and frankly terrifying) replication crisis in psychology, medicine, and social sciences highlights just how shaky the foundations are. This challenge is multiplied for cross-disciplinary studies, which wrestle with whole different ways of deciding what is true enough to act on from the incompatible ways of knowing.

Different cultures, industries, and organizations all have different standards for what count as valid interpretations to guide decisions. These standards are inevitably social — by definition, they have to be. They are about a group of people aligning on what interpretations are worthy enough to be called used for that group’s purpose, and there can’t be a logically perfect answer to that.

Opportunity from changing what we take to be true

In business, Jeff Bezos has demonstrated that “true enough” is malleable; not only do standards of truth change, but they’re also possible to change on purpose. When Amazon established Bias for Action as a leadership principle, it was correcting what Bezos saw as too strict a standard of truth. The hypothesis, apparently born out by the company’s subsequent successes, was that the desire for excessively high levels of certainty was in direct tension with reaction speed, and this was having a negative business impact. Amazon called on on workers to act when they have about 70% of the information they wish they had, an explicit call to loosen the definition of “true enough” in the interest of speed.

Across the board, tech orgs are absolutely still experiencing upheavals from changes in what they take as true enough to act on, separately and as an industry. Take user experience research. The ongoing revolution from UX has been to add validity to knowledge gained from potential customers, which could then be taken as legitimate to use in decision-making. It was mindblowing when it came into our industry and addressed a general discomfort about insular knowledge networks that lots of people had felt but hadn’t put into words.

Now, UX has established a set of practices which balance predictive power with an ability to fit into existing corporate structures (well articulated by Erika Hall in Just Enough Research). Academic-level research would be too high cost and hard to consume, but without any research, businesses would miss major opportunities. UX goldilocksed it right down the middle. Whole industries are still evolving how they judge what’s true enough to act on, and given how fast it is still changing, we have every reason to believe that there’s plenty more revolution to come.

The real work

Sure, Jeff Bezos was able to change the very definition of truth, but of course, neither of us are Bezos. It’s tempting to think of determining truth as a god-like power, sent down on tablets from a Bezos or an Ideo, not an option for us normal people. Indeed, even from our deities, these changes may feel like they wash over us, mostly leaving us unchanged. Most of us — and I certainly include myself in this — relax into existing work schemas. We’re at work, not in some epic battle to redefine truth.

But again, what if… What if we put all our various knowledge tasks under the microscope: all the little choices we make to generate ideas, narrow choices, combine, reframe, highlight, focus, and decide? Under that lens, we can see that we are deciding “true enough to act on” all the time.

If we recognize the ubiquity of knowledge choices, we open up so many new possibilities for manifesting intentionality in our work, it’s hard to take them all in. We have a constant stream of options of what to prioritize and where to draw attention. The aggregate dynamics from these localized choices is itself truth-making, not at a Bezos scale, but real nonetheless.

We have a choice whether to back off from our conflicting views and needs, living in a grey goo of agreeable compromise, or to embrace them and integrate our perspectives and needs into something new. Here’s where we have the chance to make something better than anything that existed before.

Too often this work — the real work — has to fit in the margins of work systems designed for control and production. Our industry suffers from a deep association of work with structured productive toil, a framing that’s in every way a bad fit for knowledge work. Knowledge work is uncertain and messy (and sometimes enjoyable too). The messiness can be avoided, but only at the cost of sacrificing the power and dignity of the work itself.

The other common misconception that needs addressing is that thinking is necessarily solitary — “a man thinking or working is always alone,” as Thoreau put it. This characterization just isn’t helpful if you want to improve how you think about thinking. As we saw above, deciding “true enough to work with” and improving judgement over time are dynamic social processes.

It’s a little bit scary, the idea that thinking is outside your individual control, but the evidence points in that direction. We rely on spoken and unspoken cues to navigate group dynamics, our beliefs simultaneously affecting and being affected by a group. As Mary Parker Follett, herald of modern management, put it:

“[I]n the very process of meeting, by the very process of meeting, we both become something different.” -Mary Parker Follett, Creative Experience (1924)

The tech industry suffers from a deep association of work with individual productive toil, and that just isn’t what knowledge work is. In addition to being social, knowledge work is also uncertain and messy. Moreover, it’s hard to make knowledge production happen reliably. There are better and worse ways, but in general, meaning-making argues with agendas and likes to break through timeboxes.⁴

Knowledge-making is unruly. You can’t know up-front which things are tangents. Even one distracted participant can sour a batch. Forming new knowledge requires you to be open to whole people, not just the facet of their expected role. It can also be ephemeral; if you don’t work to keep them, good thoughts vanish again. As Samo Burja says, new knowledge is often produced in small, intellectually illegitimate spaces. If we want to get better at it, we need — individually and collectively — to be comfortable with necessary mess and improve over time with social reflective practice.

Suggestions for shifting to a culture of meaning

If you’ve made it this far, I’m guessing you already basically agreed with the importance of shared meaning-making before ever reading this post. You probably also engage in productive knowledge spaces in the form of working sessions, pair or ensemble programming, or facilitated workshops.

My advice, therefore, is not about the first-order “how to make knowledge together,” but rather about how we can promote culture change to appropriately value social meaning-making and reflective practice.

My suggestions:

1. Recognize that it takes active resistance for interpretation spaces to thrive in tech orgs. There is such a strong pull in our work systems to decompose work that it’s borderline revolutionary to center integration and meaning instead. It is likely to be uncomfortable and probably won’t be directly rewarded.
2. Notice the conditions under which you are a not good as a knowledge-maker (e.g. early mornings, after too many other meetings). Avoid collaboration under those circumstances, because it reinforces the devaluation of collaborative spaces.
3. Positively reinforce the meaning-making parts in meeting spaces. Example: “this conversation was valuable — I’d hadn’t realized we were using different definitions!”
4. Negatively reinforce the one-way or mechanical tasks in shared spaces Examples: shift simpler communciations to written or recorded media, ringfence reading time for others if necessary, and most importantly, lead by actually reading what others communicate to you async.
5. Be mindful of zoom levels and different cognitive tasks. Although people may be passively gathering signals on different levels, it’s difficult and disorienting when the actual conversation shifts levels and lenses too fast. It helps to develop a richer understanding of different sub-activities of knowledge production.
6. Resist the urge to fill in the hunger for meaning with lists and tasks. Conversations with these as a primary focus are the empty calories of a working life. Fair play if they exist to provoke the useful conversations, though.
7. Shift personal development away from project management mechanics like backlog management or forecasting and towards knowledge and communication topics like facilitation, collaborative decision-making, effective writing.
8. Get inspired by learning collectives. There are many, but I’ll highlight Decolonial Futures as doing especially interesting and radical work.

And finally please. Please. PLEASE stop saying that meetings aren’t work. That’s toxic productivity culture speaking. That’s the dominant power structure trying to preserve a split between thinkers and doers. It’s a perspective that rejects the dignity of knowledge work and of your colleagues as knowers. If we want to do justice to ourselves and our fellow thinkers, if we want fulfil the promise of knowledge work as creating and not merely extracting value, we can start by honoring our meeting spaces.