Explanation may perhaps best be considered as telling a story

We tend to assume that the science and art of detection is a matter of thought and action, with explanation as a subsequent activity after the case is solved. The corollary to this in scientific research is the naive student completing his/her research and then writing up the report. It is the writing of the report and the telling of the story that is the guide in the solution of the problem. Working hypotheses, during the solving of the case, could be considered as rough drafts of the final story.

In Chapter 2 - Deduction, I considered the various metaphors used by Watson - unravelling a ball of twine, forging a chain of links, playing a game of cards - but concluded that the best metaphor for the process may be that of writing the true story.1 Most of the little cases, as expounded by Watson, start with the narration of a client, which states the problem, and ends with a narration by me, which states the solution. The narration of the client is necessarily incomplete or inaccurate - otherwise there would be no case. Watson's special genius - and I must acknowledge his skill - is in interpolating various drafts between that of the client and that of me to entice the reader to continue reading. What I had dismissed as meretricious was merely heuristic:

--- these little sketches of yours, which is entirely meretricious, depending as it does upon your retaining in your own hands some factors in the problem which are never imparted to the reader [CROO].
Alas, that I should have to show my hand so when I tell my own story! It was by concealing such links in the chain that Watson was enabled to produce his meretricious finales [BLAN].
His playing with the chronology of events in the story was also part of this heuristic device. Once again, I was unfairly prescriptive:
I am getting into your involved habit, Watson, of telling a story backward [THOR].
Just as in thinking, one can employ the technique of arguing backwards from effects to causes, as I often argued, so in writing one can employ the technique of telling the story backwards:
--- in the investigations which you have chronicled under the names of 'Study in Scarlet' and of 'Sign of Four', we have been compelled to reason backwards from effects to causes [CARD].
--- the curious analytical reasoning from effects to causes, by which I succeeded in unravelling it [SIGN].
The matter is certainly obscure. If I can't find what they are after, I must approach the matter from the other end and try to get at the principle [3GAB].
Detection is the reconstruction of a story. I have often been critical of Watson for reducing what should have been a course on the art of deduction to the telling of tales. However, on contemplation, I realize that, in a sense, he is right. Deduction is best considered as the skill in getting the story right. The detective is like a journalist who digs deeper into his stories. In each case, something happened. There was a sequence of events. My function was to tell the story.

Telling the story helps in the solution of the crime. As you tell the story , you clarify the case:

At least I have got a grip of the essential facts of the case. I shall enumerate them to you, for nothing clears up a case so much as stating it to another person, and I can hardly expect your cooperation if I do not show you the position from which we start [SILV].
Understanding is explaining to yourself. It is aided by explaining to others and eavesdropping on your own explanation. Explanation refers to a product as well as a process. The product of detection - whose art has been described in those eight chapters - is an explanation. That is, it is the story of the relevant events of the case.

1   Postmodern thinkers have argued that the stories told by scientists have no more "authority" than the stories told by anyone else. We can not prove the existence of an objective world but merely provide different subjective maps of it. There are no absolutes, but merely different statements relative to the point of view of the various observers.