The latest issue of Paranormal Review (Issue 68, October 2013), a journal put out by the Society for Psychical Research, features an article by Carlos Alvarado on the late parapsychologist William G. Roll. The article, "Attending to the Past: William G. Roll and the Old Psychical Research Literature," is not available online as far as I know.
One of the points made by Alvarado is that Roll, unlike many of his colleagues, was conversant with past research and often cited it in his work. In his groundbreaking study of poltergeist activity, Roll looked into cases dating back to the 19th century. In his discussion of apparitions, he considered the ideas of F.W.H. Myers and other pioneering researchers. Alvarado laments that many of Roll's contemporaries rarely cite the work of earlier generations, and in many cases seem unaware of it. He writes:
I would like to mention an aspect of Roll's writings that has been neglected.
I am referring to his frequent reminder of the value of the old psychical research literature via his numerous citations and discussions of these publications. In this he was following the long-established tradition in science for researchers to use references about past developments to establish continuity and discontinuity in arguments, and to seek support to present new ideas and hypotheses. But not all of Roll's contemporaries were as interested in the old literature as he was ...
Roll's attention to the old literature while he was helping us craft the modern canon should be remembered with appreciation, particularly in these days of myopic citation practices.
This has long been a sore point for me. A vast amount of work, much of it of high quality, is being overlooked by modern parapsychologists, who seem intent on monotonously reinventing the wheel. Writers like Michael Tymn, Deborah Blum, and Stephen E. Braude have made efforts to keep awareness of this early work alive, but it seems that many researchers are interested only in observations and experiments performed under current conditions. There is, it seems, an unspoken assumption that the older work has been somehow discredited by the passage of time alone.
This would be understandable if the work in question was flawed, and some of it is; William Crooks' experiments with Florence Cook, for instance, are too sketchily described to be of more than anecdotal interest today. But a good deal of the early work is both competently designed and adequately described. Examples include the Naples sittings with Eusapia Paladino, Gurney and Myers' surveys of apparitions, Hodgson's studies of Leonora Piper, Charles Drayton Thomas' work with Gladys Osborne Leonard, and Crooks' experiments with D.D. Home. Beyond these well-known cases, there are scores of others that remain of real interest; Michael Tymn's blog and books are probably the best modern source of information on them. See his book The Articulate Dead for many striking cases that are much less well-known than they deserve to be.
In some cases, strong leads from the past have petered out, unexplored. Many early experiments showed that hypnosis could markedly improve psi ability even in untrained test subjects. It's true that one celebrated case - that of the Creery sisters - turned out to be fraudulent (at least in its later stages), but this by no means invalidates all the other work. Yet the connection between hypnosis and psi is seldom examined today.
And then there's the purely theoretical side of things. Few people have ever been so deeply immersed in the subject, or have thought about it so seriously, as the pioneers of parapsychology. The hypotheses of thinkers like Myers, Hodgson, and Eleanor Sidgwick (to name three) remain deeply relevant to psi, yet are considered too rarely in today's literature.
Other sciences advance by building on the work of those who came before. I realize that parapsychology is badly underfunded and understaffed (to put it mildly), and there's a limit to what can be attempted. But this very limitation is all the more reason not to repeat work that has already been done, and done competently. Part of the motive for repeating the same kinds of experiments is, I suspect, the desire to convince skeptics by continually refining precautions against fraud. But at a certain point, this becomes an exercise in futility, since for the more dogmatic skeptics, no controls will ever be good enough. A more fruitful approach, especially given parapsychology's constrained resources, might be to accept the best of the earlier work as a given, and move on.