The approach was disconcerting—not because it was completely divorced from normal procedure, but because it wasn’t. Klemper’s offense seemed to Gurney not a matter of kind but of degree. The notion that a good detective always proceeds via pure logic and an open mind to objective conclusions concerning the nature of the crime and the identity of the perpetrator is at best a pleasant fantasy. In the real world of crime and punishment—as in all human endeavors—objectivity is an illusion. Survival itself demands that we leap to conclusions. Crucial action is always based on partial evidence. The hunter who demands a zoologist’s affidavit that the deer in his sights is truly a deer will soon starve. The jungle dweller who counts all the tiger’s stripes before deciding to retreat will be killed and eaten. The genes that urge certainty tend not to be passed into the next generation.

In the real world, we must connect the few dots we have and guess at a pattern that makes workable sense. It’s an imperfect system. So is life itself. The danger arises not so much from the scarcity of dots as from the unconscious personal agenda that prioritizes certain dots over others, an agenda that wants the pattern to look a certain way. Our perceptions of events are warped more by the power of our emotions than by the weakness of our data.

In this light, the situation was simple. Klemper wanted Kay to be guilty and therefore came to believe that she was. Dots that didn’t fit the pattern were devalued or ignored. Rules that impeded a “righteous” ending were similarly devalued or ignored.

But there was another way of looking at it.

Since the process of moving to a conclusion on the basis of incomplete data was natural and necessary, the common warning against doing so really amounted to no more than a warning not to leap to the wrong conclusion. The truth was that any conclusion might be premature. The final verdict on the validity of the leap would be rendered by the validity of the result.

That thought raised a disquieting possibility.

Suppose Klemper’s conclusion was correct.

Suppose the hate-filled Klemper had arrived at the truth. Suppose his sloppy procedures and possible felonies constituted a rotten route to the right end. Suppose Kay Spalter was, in fact, guilty of murdering her husband. Gurney had no great appetite for helping to free a stone-cold killer, no matter how deeply flawed her trial may have been.

And there was yet another possibility. Suppose Klemper’s hell-bent determination to have Kay put away had nothing to do with limited perceptions or faulty conclusions. Suppose it was a cynical and corrupt effort, bought and paid for by a third party who wanted the case closed as quickly as possible.

Suppose, suppose, suppose. Gurney was finding the echo irritating and unproductive—and the need for more facts compelling.