
What the organization learns not to say
Asking for solutions before hearing a problem can leave valuable information out. Detecting an anomaly, investigating it, and resolving it require different capabilities and authority.
Asking people to bring a proposal along with a problem has an obvious appeal. Whoever raises it has to organize what they know, consider alternatives, and take on a share of the work. For the person hearing the concern, that preparation makes the conversation easier and reduces the risk of having every difficulty passed on to them when someone else could resolve it. Within a familiar area, with enough information and autonomy, the requirement can be reasonable.
The situation changes when what someone detects is beyond their ability to resolve. They may recognize a coordination failure between departments without knowing what happens on the other side. They may notice a consequence of a decision without having the authority to change it. If a problem is heard only once a solution exists, some information is held back precisely where the means to complete it are missing. The question is how far that requirement helps prepare the conversation, and at what point it begins to prevent it.
A condition for speaking up
In their research on implicit voice theories at work (opens in a new tab), James Detert and Amy Edmondson identified beliefs people use to judge when speaking up to someone in authority is risky or inappropriate. These explicitly include the need to have solid data or solutions before raising a concern. Published in 2011, the paper brings together four studies and finds that these beliefs help explain silence beyond other individual and contextual factors.
The research also notes that these beliefs may predate the relationship with the current manager. Reluctance to speak up therefore does not, on its own, establish that this manager has punished someone for doing so. Nor does the study demonstrate that every request for a proposed solution produces silence. Its contribution is more precise: it allows us to examine a rule of self-censorship that can look like a norm of good professional conduct.
Applied to management, this reading calls for a review of what is actually being asked. Preparing an observation makes it easier to discuss. Requiring its author to answer every question before presenting it turns that preparation into an entry requirement that is difficult to meet. The filter favors problems one person can understand and solve; those spanning several areas of responsibility may have no one to hear them.
Information is distributed before there is a solution
Asking them to specify the cases, what they observed, and why it concerns them helps distinguish a recurring issue from an impression. Asking for a complete solution before hearing them out adds a different burden: resolving, from a single role, a problem whose diagnosis requires input from several roles. The organization may wait for a proposal that this person is not in a position to produce. Meanwhile, the requests keep circulating.
Differences in position already matter when disagreement contains information about what each person was able to observe. Here, the difficulty arises before perspectives can be compared: some may never be expressed. Recognizing a local pattern can be enough of a contribution to begin connecting it with other observations. Describing that scope honestly is already a contribution; bringing the pieces together may be a separate responsibility.

The evidence needed depends on the next step
There are intermediate decisions between hearing a warning and changing a process. A record can be checked, another department consulted, or an exception examined without yet accepting an explanation. Each step has a cost and a scope. Requiring the same certainty to begin that review as would be needed to intervene leaves no path forward for signals that can only become more substantiated once shared.
This does not turn every concern into a diagnosis. A warning needs some basis that can be examined, even if it is initially partial: an identifiable episode, a difference from what was expected, or a consequence that deserves attention. The person communicating it must distinguish what they observed from what they assume. The response can ask for greater precision without making a solution a condition for being heard. It can also conclude that the information provided does not justify further investigation, as long as it explains what was reviewed and why.
The requirement should match the immediate decision. In the case of the requests, checking whether there is a recurring issue requires less evidence than redesigning the workflow between departments. A limited check may end up ruling out the hypothesis. It may also reveal that the problem requires someone with access to the system or authority over the process. If the person who first signaled a problem is expected to complete all of that in advance, a lack of information becomes a barrier to obtaining it.
The response defines what is worth saying again
A general invitation to participate leaves a practical question open: what happens when someone brings a concern they do not yet know how to resolve? How the work is distributed from that point matters. If information from another department is needed, someone with access must help obtain it. If investigation is required, it must be clear who can take it on. The person who detected the anomaly can contribute without automatically becoming solely responsible for resolving it.
That distribution allows a meaningful standard to be maintained. Whoever flags a problem is responsible for the quality and limits of their observation. Whoever coordinates the review is responsible for ensuring a proportionate response. Neither responsibility requires a promise that something will change, but both need something more concrete than thanking someone for the comment and leaving it pending with no way forward.
Organizational dark matter raises the question of which dynamics are left out of the model used to make decisions. This barrier operates at the point of entry into that model: what must a person already have resolved for their information to be considered? The absence of objections does not tell us how much was left out. The rule governing how those that do emerge are received is worth examining. When communicating an anomaly also requires mastery of its solution, signals about problems that can only be understood by bringing together several people’s knowledge may be left out.





