Last week’s column set some readers’ teeth on edge (“Why Johnny Corporation can’t learn,” Keep the Joint Running, 11/19/2007). The issue was its negative tone.

The column explained why corporations rarely learn from their mistakes and even more rarely learn from their successes. “Don’t bring us problems — bring us solutions,” was the gist of the complaints.

Fair enough, so long as we all agree that unless you understand the underlying dynamics of a problem you’re unlikely to find a solution.

Solutions to these two problems are especially tough to implement because doing so calls for challenging some very popular business beliefs and practices. In particular, companies that truly want to be “learning organizations” will have to thoroughly and deeply re-think their approach to accountability, to metrics, and to compensation.

The first, and possibly most important step you can take is to give up one of the most cherished notions in American pop business culture — the need to hold people accountable. KJR has covered this subject before (see “The dark side of accountability,” 11/3/2003).

Among the many disadvantages of a hold-people-accountable business culture is that it pretty much guarantees that anyone associated with a mistake will do everything possible to hide it — to sweep it under the rug, bury it in a quarry, or put sawdust in the crankcase and call it a mint-condition classic.

Or, if there’s no way to hide the mistake, they’ll find some poor schnook to serve as a scapegoat. Root cause analysis? The root cause is that this schnook here caused the problem.

Think anyone will admit to a mistake under these circumstances? Of course not.

So Lesson One in creating a learning organization is to create a culture of responsibility. That means consistently praising everyone who identifies mistakes and takes the lead in fixing them, without ever paying the least bit of attention to who made the mistake in the first place.

The next place to look is metrics. Last week’s column pointed out that when your metrics define failure as success and vice versa, they guarantee you’ll repeat and extend your mistakes because you always get what you measure.

The best metrics assess your progress toward real, meaningful goals. They aren’t proxies or indexes, and they don’t assess intermediate results.

To illustrate, imagine three police departments. The first measures success by the ratio of arrests it makes to crimes committed. The crime rate goes up, not down, because the city’s peace officers are insufficiently concerned about determining who actually committed each crime, and even less concerned about making sure the evidence they present isn’t thrown out because of irregularities.

The second police department, headed by wiser leaders, measures the ratio of convictions to crimes. Much better. For the most part, its officers take care to arrest only those who seem to be guilty.

Much better, but still not as good as the third police department, which measures its success in terms of overall public safety.

Which is why Lesson Two in creating a learning organization is to choose your goals wisely and measure them carefully. Doing so points you in the right direction and makes sure you know success and failure when you see them.

Lesson Three is simply stated but difficult to put into practice: Align compensation. It’s particularly important if you want to learn from success.

This isn’t a matter of rewards and motivation, because compensation, properly understood, isn’t about rewards and motivation. It’s about the most tangible form of communication a company has — what it pays for. To understand this point, compare these two phrases: “Here’s what we want you to do,” and “Here’s what we’re paying you to do.”

The latter packs a much stronger punch, don’t you think?

Aligning compensation means defining acceptable performance as delivering results and strong performance as delivering repeatable results. Exceptional performance? That means repeating the results of others, not having them repeat yours.

Whether the subject is reusable software modules, successful business practices, or borrowing another employee’s PowerPoint deck, if you want a company that learns from its successes you had better provide the highest compensation to those employees who find those successes and learn from them.

Even this is dangerous, by the way. Because if you aren’t careful you’ll stifle innovation along the way.

It has to be that way. Learning from success means you don’t reinvent the wheel.

Innovation means you do.