One of the characteristics I’ve noticed quite a bit in detective stories from the golden age of mysteries (roughly, From 1890 until the start of World War 2) is how many detectives had a theory of detection which they discussed.
In the very early days, the detectives differentiated themselves from the police through their use of forensic investigation. In the 1890s, Sherlock Holmes performed chemical analysis to prove a stain was blood and wrote a monograph on how to identify cigar ash. In the early 1900s, Dr. Thorndyke looked at everything he could under a microscope, and what he couldn’t he would look at with enlarged photographs.
Sherlock Holmes did not long predate real forensics, though. By 1901 Scotland Yard was using fingerprints to identify people and in 1902 the first conviction was obtained with the use of fingerprint evidence. (See Fingerprints And Forensic Evidence.) It did not take the police long to make use of this kind of forensic evidence, and private detectives began to shift their methods. G.K. Chesterton would revolutionize the field of private detection in 1910 with Father Brown’s psychological approach to solving crimes, and to varying degrees this has been the primary tool of detectives ever since, so no advances in forensic technology can make psychology obsolete.
Through all of these changes, there remained an air of novelty. The brilliant detective during the golden age was not merely brilliant; he had a method. He got his results because he brilliantly followed his method while others either followed the wrong method or else had no method.
Perhaps nothing exemplifies this as well as the unjustly neglected second Poirot novel, Murder On the Links. Poirot’s method is contrasted very strongly with that of the indomitably forensic M. Giraud. Giraud examines the crime scene with the utmost care and uncovers impressively small clues. Yet Giraud dismisses a section of pipe as being of no importance because it’s not the kind of clue he’s looking for. As Poirot remarks to Hastings, “Mon Ami, a clue of two feet long is every bit as valuable as one measuring two millimetres! But it is the romantic idea that all important clues must be infinitesimal!”
Poirot considers all clues because his method is to adjust his theory until nothing is out of place; Giraud’s method is to ignore whatever does not fit his preferred kind of evidence. The point, here, is not the specifics of the contrast, but that the contrast is so important.
Another, though less important, example that comes to mind is in The Strange Case of Mr. Challoner. In it the detective, Malcolm Sage, delivers a lecture on how the Police misunderstand the evidentiary value of photographs and fingerprints. They think that photographs are for identification while fingerprints are evidence; in reality fingerprints are for identification and photographs are evidence. He took a series of photographs of the crime scene and announces that they will be the principle evidence at trial, and then uses fingerprint evidence to show that the butler is actually a wanted criminal. I don’t know that the police ever ignored the identificative value of fingerprints or the evidentiary value of photographs, but that’s not the point. In a short story written for entertainment value, the writer and editor thought that the audience of the newspaper would be entertained by a lecture on how the police don’t understand the proper use of evidence.
I’m not sure exactly when this aspect of detective fiction died off. Certainly you can’t find it in the Cadfael series, which started in the 1970s. I can’t think of any detective fiction I’ve read from the 1940s through the 1960s except for Miss Marple. I haven’t read any of the Poirot stories written after 1947 (yet). I don’t remember this in the Miss Marple stories from that time period, but then I don’t recall it in the Miss Marple stories from the golden age, either. (To be fair, that’s only one novel, though it is also most of the short stories.) Miss Marple was never really a detective, though. People told her things and then she would give them the solution. With the exception of Nemesis, and to a lesser extent A Caribbean Mystery, she never went looking for clues of any kind. On the other hand, there were her typical reminiscences of people who committed similar sins in Saint Mary Meade, which was certainly a unique style of detection.
By the time we get to television detectives like Columbo in 1971, the aspect of a unique method is missing. While it might be objected that Columbo is a policeman and therefore cannot contrast with policemen, he is still a contrast with the other officers who do not get nearly the same results.
There is similarly no trace of in the 1980s’ Murder, She Wrote.
So, what happened?
Alternatively, what was special about the golden age?
I’m really not sure which of these questions we should be asking. It is tempting to think that there was something special about the time that the golden age happened. To some degree it was the first time police forces were getting organized and police detectives were becoming a real thing. Advances in technology also made various kinds of detection newly possible, or at least newly practical, and so the whole thing had an air of novelty to it.
On the other hand, it’s also possible that there was simply a fundamental split in the mystery genre, with mysteries taking the psychological and logical aspects of detection and police procedurals taking the forensic aspects of detection.
On the third hand, it may just be that all of the possible theories of detection have been expostulated and all that remains is to do one of them well.
Perhaps it’s a bit of all three.