Cognition problems

Cognition problems are questions for which a definitive answer exists, or for which some answers are better than others. They are one of the three broad kinds of problems through which the book examines collective intelligence, and they show why the crowd’s potential depends not just on gathering people but on structuring how their knowledge contributes to a judgment.

Definition and Examples

Some cognition problems have plainly verifiable outcomes: who will win the Super Bowl, how many copies of a new ink-jet printer will sell in the next three months, or how likely a drug is to receive approval from the FDA. Others lack one uniquely correct answer but still admit better and worse judgments. Choosing the best place to build a public swimming pool is such a question: the decision can be assessed, even if it cannot be settled by a single factual result.

This range distinguishes cognition problems from questions that primarily require people to coordinate their behavior or cooperate despite conflicting interests. The book treats cognition as a broad category, encompassing both forecasts with eventual outcomes and practical judgments whose quality can be compared.

Place in the Book’s Argument

The book organizes its first half around cognition, coordination, and cooperation, alongside chapters on the conditions that make collective intelligence possible: diversity, independence, and a particular kind of decentralization. Cognition problems therefore introduce one part of a larger inquiry into how groups can reach sound decisions—and why they sometimes fail. The case studies in the second half extend that inquiry to different ways of organizing people around common or loosely shared goals, examining where collective intelligence flourishes and where it flounders.

The category also clarifies the book’s central distinction between possessing information and making it collectively useful. Decentralization encourages independence and specialization, allowing people to develop local knowledge and solve problems close to where they arise. But information uncovered in one part of a system may never reach the rest. The challenge is to preserve the value of specialized knowledge while creating a means to aggregate it into a collective judgment.

Aggregation and Organizational Failure

The pre-September 11 U.S. intelligence community illustrates the gap between distributed expertise and effective collective cognition. Its agencies’ division of labor fostered specialization and varied analytical perspectives, but the system lacked a real mechanism for aggregating judgments as well as information. The result was decentralization without aggregation: potentially valuable knowledge remained divided among agencies rather than becoming a shared assessment. The proposed remedy is not simply to place a central authority over the system, but to enable aggregation without discarding the diversity and local knowledge that decentralization provides.

A different failure appears in NASA’s handling of the foam strike on the Space Shuttle Columbia. The Debris Assessment Team engineers had serious concerns and said that images of the Orbiter’s wing were needed for an informed analysis, yet the Mission Management Team did not give those concerns a serious hearing. Its meeting practices narrowed the range of analysis: discussion largely took the form of a question from the leader followed by one answer, without a deliberate effort to hear from other members. The team also began with a conclusion and reinterpreted incoming information to fit it. This episode shows how a group can possess relevant expertise yet fail to turn it into a sound collective judgment.

Information Flow and Local Knowledge

Rigid corporate hierarchies create a related obstacle. At Ford in the 1970s, more than fifteen management layers separated the chairman from a factory-floor supervisor; at Toyota, the equivalent distance was five. In one General Motors plant in Van Nuys, California, two workers had spent two years welding a bracket without knowing that it held the car’s sunshade. When a supervisor showed them the downstream repair required if they missed it, one worker was shocked to learn what the part was for. The example makes the cost of severed information flows concrete: workers may perform a task without access to knowledge that could help them understand or improve it.

The same hierarchy could make even the path of an idea through a company forbiddingly long. A chart reproduced in In Search of Excellence from an unnamed company showed 223 possible paths through its bureaucracy for a new product idea to be accepted. Layers obstructed not only upward communication but also executives’ ability to know whether their picture of the organization matched conditions on the ground. Cognition problems are therefore not confined to formal forecasting: the capacity to make good organizational judgments also depends on whether relevant knowledge can travel through the system.

The book’s proposed organizational principle is to let people close to a problem make local decisions whenever possible. Experience gives them tacit knowledge that is difficult to summarize or convey, while specialization can be more valuable than managerial expertise in choosing a workable solution. Yet decentralization alone is not enough. A capable group must preserve local judgment and independence while connecting specialized knowledge so that it can inform decisions beyond the place where it originated.