Six Ways
Communication Failure
Is Misattributed.

How communication shows up in organizational culture

If you have been following my frameworks on my website, you know that communication architecture starts with Ed Schein’s organizational culture pyramid. The premise is that culture is measured by the alignment between artifacts, espoused values, and tacit assumptions.

A change to an organization’s communication architecture starts with pulling the tacit assumptions back into line with the values you espouse. To do that, there are six failure dimensions that are commonly misattributed to the person who communicates differently.

Where you file communication breakdowns today

Think about where a communication breakdown actually gets recorded today in your organization. It is on an employee file, like a performance review for example. And in this review, it is considered a pure soft skill that is solely assumed to be the individual’s effort that drives success. So it becomes a development area, which becomes a coaching engagement or a workshop, and repeats performance cycle to performance cycle. Anyone who has been a leader has written these reviews and seen the same in their own reviews, because leadership is influence and influence is communication. It is meaningful feedback that is important for being an effective leader.

I argued last week that the failure is not purely the individual’s effort. It happens at what I call the interface, in the space between how one person thinks and how the room listens. If you accept that premise, you still walk into your next review with no changes to how you think about communication feedback.

That is the practical problem I want to solve for you this week. Nobody in your meetings can name what really failed, and so the proxy category is automatically assigned. And that proxy is individual development opportunity. As a result, the coaching budget you allocated gets spent as development. The risk analysis that was critical never reaches the decision-making meeting. And the decision that the accountable leader needed to own and make does not get made.

The reframe is articulated in my Cognitive Translation Protocol. It names the six points where communication is misattributed and affects judgment reaching decision makers in organizations. Each fails for its own reason, and each is specific enough to design against. Below, every point gets a moment you have sat in, and one question to ask in your meetings. The repairs for each come later in this arc so today is diagnostic understanding only.

Take a breakdown that has already recurred in your organization, ideally one with a person’s name attached. Now carry that breakdown through the six dimensions as you read. What I want you to be able to do by the end is say which of the six it is.

Where the six sit beside the training you buy

A leader who hears that there are six failure dimensions will place them next to the communication training the organization already pays for, so let me put the two side by side before the six arrive.

Nearly all of that training works on the sender. Presentation skills, message architecture, storytelling, executive presence, media coaching: your budget goes into shaping the message better and delivering it better, on the working assumption that the people receiving it assemble meaning about the same way you do. Where that assumption holds, the training pays in full. The six dimensions examine the assumption itself.

Your organization already accepts one large exception to that assumption, and has been funding the exception for decades. Every global leadership program teaches the difference between high-context and low-context communication, because the same message assembles differently in Amsterdam and in Tokyo. The relaxed, first-name senior leader who reads as confident in an American room can read as unserious in a Japanese one. This is national culture rather than Schein’s organizational culture, and the reason it belongs here is what your organization does about it. You do not send that leader to coaching. You brief both sides and you change the format, because you understand the mismatch as sitting between two contexts rather than inside either room.

The same effect now shows up in machines, which is worth a minute because of how deep it suggests the principle runs. In July 2026 Anthropic published an analysis of roughly 310,000 conversations across three of its models and twenty languages, and found that Claude expresses different values depending on the language the conversation is held in: the most warmth in Hindi and Arabic, the most rigor in English and Russian. Two people asking for feedback on the same business plan in different languages can come away with different impressions of the plan. That is meaning assembly travelling with context, in a system trained on most of the written record.

The six dimensions ask you to carry a principle you already accept one level further down, from context to cognitive system, and two things make that harder than it sounds. The variation runs inside a single office, a single team, a single meeting, so nothing warns you that the contexts changed. And your own training has taught you where to look when communication fails, which is at the message, at the delivery, and at the person. After enough years that is where you look first, without deciding to. Seeing the interface means interrupting a pattern your own development budget installed in you, which is a different ask from adding one more skill.

So hold both of these at the same time. Language and influence matter as much as your training says they do, and they will matter more once you have designed the interface, because a well-crafted message that fails at one of these six dimensions takes your craft down with it. What the six give you is the list of places where a crafted message fails, and everything below assumes you keep the craft you already have.

Inference: when a precise answer gets read as hedging

Start with the dimension you will recognize fastest. You ask a director in a steering meeting whether the timeline holds. She answers the question you asked: the build is on track, integration testing is the open risk, here is the date to watch. What you hear is hedging. By the following week the account of that meeting is that she would not commit, and at mid-year it arrives in her review as a note about executive presence.

Look at the exchange again. What she transmitted was a delivery status and a named integration risk, and what you received was a mood. The words were accurate when she said them and audible when you heard them, so the loss happened in the layer where the two of you assemble meaning. I named this dimension inference: the mismatch between what was communicated and what was inferred. Literal and implicit communication styles produce systematic gaps when the interface is not designed.

Some of the people in your meetings state what they mean and mean what they state. Others state part of it and expect you to assemble the rest from context, tone, and shared history, because in their experience that is how competent adults talk. Put those two on opposite ends of one exchange with nothing designed between them, and the literal answer keeps getting read for subtext it does not contain, while the implicit one keeps getting taken more literally than it was built to be. Anyone who has asked for a quick gut check and received twenty careful minutes has been standing at this dimension. Quick was implicit, and it landed on someone who processes literally.

The misattribution is unwillingness to commit. What actually happened is that the room supplied a meaning the sender never sent.

The audit question: where in your meetings does meaning get filled in rather than stated? What you are looking for is a delivery you keep interpreting the same way, so that the direct answer keeps arriving to you as reluctance, or the carefully qualified one keeps getting filed as doubt when it was the most accurate answer in the room.

Timing: when the considered answer arrives after the decision

You go around the table at the Tuesday leadership meeting. Three people answer on the spot, a fourth says something noncommittal, and you close the decision on the three answers that were in the room. The fourth person had the most complete answer of anyone at that table and needed until the next morning to assemble it, because assembling is how that person produces answers. Your meeting format closed the question before that answer existed, so you never got to weigh it against the three you did hear. Cognitive systems process and respond at different speeds, and environments that reward immediate response systematically disadvantage deliberate processors.

Anyone who runs a standing leadership meeting has closed a question this way, and it is worth being honest about how invisible it is from inside the room. Nothing failed. The discussion was energetic, you made a decision, your minutes record that a decision was made, and everyone moved on to the next item on the agenda. The loss shows up later if it shows up at all, as a decision that would have gone the other way, and nothing anywhere in your organization records the answer that was still being assembled when you closed the question. Your minutes have no line for it, and neither does anything else you keep.

The misattribution is that the person does not contribute in meetings. What actually happened is that your format set a response deadline the person’s processing does not meet.

The audit question: what does your meeting format do with an answer that needs a day? In most formats the honest answer is nothing. There is no slot for “I will have the real answer tomorrow,” so the record closes without it every time, as a matter of format rather than judgment.

Signal clarity: when your message assembles three different ways

You send two lines on a Tuesday: revisit the Q3 numbers before Thursday. To you that is unambiguous, because you know which numbers, which concern, and how far revisit goes. The three people who receive it spend part of their afternoon deciding what you meant. Is this a sanity check, or is the plan in question? Does before Thursday mean a reply or a rebuild? They compare notes, prepare for the heavier reading to be safe, and the cost of the specificity you left out gets paid by three people who were not in a position to supply it themselves. Seniority makes this worse rather than better, because the more senior you are, the more certain you tend to be that what was obvious to you traveled with the message. Anyone who has led at any level has sent that message.

My protocol defines signal clarity as the difference between what was meant, what was said, and what was heard. Without explicit signal design, clarity is assumed rather than built. Three versions of your message exist in every exchange: the one in your head, the one that left, and the one that landed. With nothing designed at this point, you proceed as though the three were the same message.

The misattribution is that your people cannot take direction, or that they need too much hand-holding. What actually happened is that you shipped a message with the specificity removed and no design to recover it.

The audit question: where does your organization assume clarity instead of building it? Look at the messages that reliably generate a round of clarifying questions, and the briefs everyone nods at that no two teams then execute the same way.

Sensory load: what the room costs before anyone speaks

Your quarterly all-hands runs in a bright hall with a hard echo, screens on two walls, a deck moving faster than anyone can read it, and side conversations at the edges. You put important content into that room, and only part of it reaches the people you called there. Of the six, this is the dimension I expect you to set aside first, because it does not look like communication at all. It looks like a facilities decision.

Environmental factors consume cognitive bandwidth before communication begins, and high sensory load environments reduce signal fidelity for all participants, selectively for some. The attention your people spend managing the room, filtering the noise, keeping up with the pace, is attention that never reaches your content. How much it costs varies by person. Some nervous systems register far more of the environment than others, so the same hour costs one of your people a small overhead and costs the person beside them most of the capacity they walked in with. Both of them attended, and only one had enough left to follow you.

The misattribution is disengagement, because the difference in what each person took away looks like a difference in how much they cared. What actually happened is that nobody in your organization counted what the room itself consumed.

The audit question: what is the room costing you before anyone speaks? Take your most consequential recurring meeting and account for the room itself: noise, light, pace, how many channels are running at once. Whatever the room consumes comes out of your content before your content gets a hearing.

Emotional attribution: the emotion your room reads into a tone

A manager gives you a project update in a level, even register, laying out the facts without the voice rising once, and it is done in four minutes. You come away with an impression that the manager is low energy and possibly checked out. The update itself was careful, complete, quietly proud work. What travels out of that room is your impression, into the talent review and onto the succession slate, and the quality of the work you were actually told about does not travel with it.

That is one direction, and it is the direction leaders name most easily. This dimension runs in every direction. An engineer answers a challenge directly, without softening the disagreement, and what gets recorded afterwards is that he was angry, or that he came at people, and it turns up in the 360 as abrasive. A director delivers a plan at pace and emphatically, and the note afterwards is that the delivery was emotional, or that the director is not ready for the executive team. Someone runs warm and enthusiastic in a room that treats warmth as a sign of not being serious, and the read is that the work lacks rigor. In each of these, including the flat one, the room did not evaluate what was said. The room assigned a state of mind to a delivery, and then evaluated the state of mind.

I named this dimension emotional attribution: the automatic assignment of emotional meaning to communication style. Tone, cadence, and delivery are interpreted through one cognitive default and misread when the sender uses a different one. The definition names no particular emotion, and that is deliberate. Any state your room is capable of reading into a delivery belongs here: energy, anger, defensiveness, enthusiasm, confidence, composure under pressure, investment in the work.

Your room carries a default for what engagement is supposed to sound like, and for confidence, for concern, for disagreement, and for holding your position when you are challenged. The people whose natural register matches that default have their actual state understood correctly without anyone doing any work to get there, and nobody notices that any interpreting was required. The people whose register sits away from that default get interpreted every time they speak, and what you interpreted gets written down afterwards as an observation about them. Nobody records “that tone, read through my default, suggested disengagement,” and nobody records “that pace, read through my default, suggested anger.” They record “seems disengaged” and “gets heated in reviews,” and the next person to open the file reads both as fact. I should declare my interest here, because my own register is flat, and for years the content of what I was saying rarely got checked against the emotion a room had already assigned to it.

The misattribution is a state of mind, and it can be any state of mind your room holds a default for: low energy, disengagement, anger, defensiveness, over-emotionality, arrogance, or a general absence of executive presence. What actually happened is that you read a register against your own default and recorded the reading as an observation about the person.

The audit question: whose tone gets interpreted, and whose gets taken at face value? The people who match your room’s default get received accurately without anyone working at it, everyone else gets actively interpreted, and once that interpretation is written into a talent review it is read as fact by every person who opens the file afterwards.

Execution visibility: when everyone leaves with a different decision

The last dimension fails after you have stopped watching. Your meeting ends well, the path feels clear, and two weeks later two teams have built two different things, each faithful to what its own people understood you had decided. Two people left your room with two different decisions in their notes, and nothing between that meeting and the build made the difference visible to you.

My protocol names this dimension execution visibility: the gap between what was decided and what was understood to have been decided. Without explicit execution architecture, inference fills the gap differently for different cognitive systems.

What makes this one expensive is where it sits. The other five fail during the exchange, while somebody in the room might still catch it. This one fails on the way out, in the space between “we decided” and “here is what we decided, who owns it, and by when.” A decision that leaves your room as a shared feeling of agreement actually leaves as four or five private versions of a decision, one in each person’s notes, and you do not find out how far apart those versions are until the build makes them concrete. That is the most expensive place in your organization to discover a disagreement that was sitting at your table all along.

The misattribution is poor execution, or a team that does not follow through. What actually happened is that the decision left your room without an artifact, so each person carried out the version they had assembled.

The audit question: how does a decision leave your room? If the answer is that everyone just knows, this is your dimension.

The six, and what each one gets recorded as

The Operating Brief: run the audit

Name the dimension, then design for it

Each of the six is a place where communication gets misattributed to a person, and each fails for a reason specific enough to design against. That specificity is what the names are for. “Communication is broken” leaves you with nothing to act on. “This analysis is failing at inference, in the Thursday review, because the room keeps filling in intent from a delivery it was never built to read” is a sentence you can say out loud to your leadership team and act on by Monday.

Going back to Schein, the six dimensions are where your tacit assumptions about communication sit furthest from the values you espouse about it. Naming which one is failing is what makes the two available to align.

Against the six dimensions the protocol sets nine structural interventions, in three lanes, and the next three essays take one lane each. The interpretation interventions come first: interpretation guardrails, which separate observation from interpretation; context equalization, which establishes shared understanding before judgment; and outcome-based evaluation, which reads results rather than inferred intent. Then the process interventions: signal persistence, pacing and timing controls, and decision traceability, which is the record that runs from signal to interpretation to decision. Then the environment interventions: cognitive offloading, environment abstraction, and boundary encoding, which is the explicit statement of roles, limits and stop-rules.

One thing about that structure is worth knowing before those essays arrive, because it changes how you use today’s audit. The nine interventions do not sort one to one onto the six dimensions. Each intervention addresses three dimensions, and every dimension is addressed by interventions sitting in more than one lane. Inference is addressed by six of the nine, across all three lanes. Sensory load is addressed by three, in the process and environment lanes. So naming the dimension is the diagnostic act and it tells you which interventions to reach for, and it does not point you at one lane and stop.

You have the names now. Walk your own meetings this week and locate one breakdown on the six. That is the audit, and for this week it is enough.