I've been thinking a lot about how to present my arguments. Johnson and Chappell considered all the evidence at their disposal but offered no arguments. They simply stated their conclusions. Leach judged the quality of their evidence to be poor, offered some scraps from Jamaica, and allowed that Jamaica might have been the John Henry site. Nelson suffered extreme confirmation bias by considering only evidence that, in his view, supported John Henry at Lewis Tunnel in Virginia. I think that all of these studies are seriously flawed, and I don't want to emulate any of them.

One obvious logical principle is that *all* of the available evidence must be considered. There is a great deal of it, so this quickly leads to a complicated presentation, even though the final conclusion may be clear cut.

I think I've found a short cut. For WV, Jamaica, or VA, consider only the evidence that supports that site. In other words, construct the best-case scenario for each site. Perhaps the evidence does not discriminate between that site and others, even in the best-case scenario. If this were true, then WV, Jamaica, and VA would not merit further consideration. I believe that it *is* true.

WV: C. S. "Neal" Miller testified that he witnessed John Henry's contest with a steam drill at Big Bend Tunnel in 1870. Census records show that he did not live near Big Bend Tunnel in 1870, that he was 7-8 years old, and that he attended school. Miller's testimony is a fabrication. The testimonies of other men who claimed to have worked on Big Bend Tunnel during its construction are indirect and highly contradictory. About 40% of versions of the ballad "John Henry" published by 1933 place JH at Big Bend Tunnel, on the C & O RR, or both. Steam drills were not used in boring Big Bend Tunnel, but one could have been brought there for a trial, part of which could have been a contest with JH.

Is this evidence plausible, given that JH *was* at Big Bend Tunnel?
Yes.
Is this evidence plausible, given that JH was *not* at Big Bend Tunnel?
Yes. Contradictory testimonies do not depend on whether or not JH was at Big Bend Tunnel. Relocalization is so common in folksong that places they give are not reliable.

The evidence that is alleged to support the hypothesis that John Henry was at Big Bend Tunnel, Summers County, WV, does not discriminate between that hypothesis and others.

Jamaica: A few song fragments mention John Henry, and there is testimony that he died while working there on a road or railroad in 1894-96.

Is this evidence plausible, given that JH *was* in Jamaica?
Yes.
Is this evidence plausible, given that JH was *not* in Jamaica?
Yes. The tradition could have been taken from the US to Jamaica by laborers. Indeed, this is highly likely. Further, there is testimony that the ballad "John Henry" was known before 1894.

The evidence that is alleged to support the hypothesis that John Henry was in Jamaica does not discriminate between that hypothesis and others.

VA: A John William Henry was a convict in the Virginia Penitentiary who was leased to work on the C & O RR. He then disappeared from the records. A stanza of the ballad "John Henry," found occasionally, states that he was taken to the "white house" and buried "in the sand" where locomotives pass by. At Virginia Penitentiary there was a white workhouse and a mass grave with a nearby railroad. Steam drills were used in boring the C & O's Lewis Tunnel in Virginia. The hypothesis is that John William Henry was a steel driver who raced a steam drill and died at Lewis Tunnel, thereby giving rise to the legend. No song or legend places JH at Lewis Tunnel.

Is this evidence plausible, given that JH *was* a steel driver who raced a steam drill at Lewis Tunnel?
Yes.
Is this evidence plausible, given that JH was *not* a steel driver who raced a steam drill at Lewis Tunnel?
Yes. Men named "John Henry," white houses, sand, and railroads are all common enough for the observed correspondences with a stanza of the ballad to be pure coincidence. There is no evidence that John W. Henry was a steel driver, that he was ever at Lewis Tunnel, or that he died there.

The evidence that is alleged to support the hypothesis that John Henry was at Lewis Tunnel in Virginia does not discriminate between that hypothesis and others.

Before any evidence is considered, all tunnels that were bored in the South between between the Civil War and about 1888, when the ballad appeared, have equal probabilities of being the John Henry tunnel. I don't know exactly how many there were, but I'm sure there were at least 100, probably many more. Thus, the "prior probability" that JH was at Big Bend or Lewis Tunnel is no greater than 0.01. Since the evidence, in the best-case scenarios, does not discriminate between these and other sites, this evidence, after being considered, leaves the probabilities of Big Bend and Lewis Tunnels at their prior value, 0.01 or less. The prior probability for Jamaica is even smaller, and it, too, is left unchanged by consideration of the evidence that is alleged to support it.

As JH sites, Big Bend Tunnel, Jamaica, and Lewis Tunnel can be dismissed for lack of evidence.

Only one site that has been seriously considered remains: a tunnel at Dunnavant, Alabama. In that case, the evidence is more complex.