(Edit: this is my analysis of the synopsis, looks like some of the formatting got screwed up in the conversion and there are spaces missing for some reason in many quotes.)
Please note that this is the only chapter of the book that I’ve read – though I probably will go through the rest of it at some point as it looks to be an interesting read and I like his basic approach ( accounting for the times).
Note that this book was written in 1886 and the author’s use of language is formal and not always easy to decipher. I’m going to lay things out as I interpret what he writes, though I expect that there may be different interpretations possible to certain passages. There are some bits where I really wasn’t certain what he was saying. I’m not holding out any of the following as definitive, but rather as an opening for discussion. Especially that I’m basically commenting as I go rather than giving a dissertation-like level of study to this text.
The chapter is titled “General Criticism of the Evidence for Spontaneous Telepathy”. In it he reviews problems with past research in this area and describes some of the methodology that he is using for this experiment along with a summary of his overall conclusion of the study.
(Note the headings I provide below are my own. I provide section references (§ x) to the actual text throughout. I provide my own thoughts and questions I ask myself in brackets as well. Note that most of the comments in brackets were put in as I read though I have gone back over it and added a few more.
The difficulty with studying spontaneous instances of telepathy
The author begins by noting the difficulty with studying possible spontaneous instances of telepathy: mainly that the person isn’t going to be aware of it coming in advance, making for difficult scientific study. He notes (§ 1.):
The methodofinquirywillnowhavetobethehistoricalmethod, andwillinvolvedifficultquestionsastothejudgmentofhuman testimony,andacomplexestimateofprobabilities.
He notes some of the difficulties with the evidence are:
· There are many people in the world.
· Mal-observation is easy and common.
· Exaggeration is easy and common.
(He’s alluding here to the law of large numbers and the falliability of human observation and tendency to exaggerate. I agree. One could add basic mis-remembering and confabulation here too – he refers to these below.)
Past Studies:
He recognizes that these are particularly problematic when it came to previous studies such as those evaluating the evidence in favour of witchcraft where:
· Direct evidence came only from the “uneducated class” (hey – his words, don’t blame me!)
· Those who studied it didn’t account for the possibility of self-deception on the part of the witnesses.
He notes that we now know that hypnotism, hysteria, and hysteron-epilepsy can account for many phenomena attributed to demonic possession. Note: this is not fraud, but genuine hallucination.
He notes that other more “bizarre and incredible marvels” had no direct, first hand, independent testimony.
The best cases he knows of in past studies could be explained as telepathic, but the evidence is not strong enough to support “definite conclusion.”
(The impression he gives is that previous work in this area is pretty fraught with risk of error.)
Present study:
He contrasts those past studies with the present study which he calls a “complete contrast” to those previous studies.
(I’ll have to take his word on this, I’m not familiar with the past studies)
The evidence in his study, he argues:
· Comes from mostly educated people, who were not predisposed to believe.
(I’m curious as to what exactly he means by “educated” and what he believes the significance is. He doesn’t elaborate much on that in this chapter. Presumably he describes this in more detail elsewhere in the book).
· The phenomena are not associated with prevalent beliefs or habits of thought.
(I’m curious about this as well as he doesn’t elaborate further here. He seems to be saying that the witnesses have no biases towards any particular view – although laudable to attempt, my view is that everyone has biases. That said, I get what he’s trying to do (assuming I’ve correctly interpreted it).
He notes despite this, we should not simply assume the evidence is trustworthy.
(He quite correctly notes that a person being educated and presumed without bias is not enough to rely on the account. When he writes “trustworthy” I believe he’s talked more about error than deliberate deception – which is my focus as well.)
He notes possible errors in the present study: (I consider this akin to risks of bias that Cochrane refers to.)
1. Errors of observation resulting in mistaken identity:
· stranger might be mistaken for a friend who turns out to have died at that time – and whose phantasm is asserted to have appeared
· he notes that this applies only to a small minority of the cases.
(I am interested to see how he evaluates this: error of observations can go beyond mistaken identity. Depends on the situation)
2. Error of inference:
· he says this is not a high risk of bias (his words are prominent danger)
· I think he’s saying here that the evidence they collected was what the percipient reports having seen or heard – not what the percipient inferred regarding what they saw or heard
(I can see this: he’s not documenting what the prescient infers about what they experienced, just what they experienced.)
3. Error of Narration:
· He considers this risk to be much higher, due to the tendency to make an account:
o edifying,
o graphic, or
o startling.
· He notes that the tendency towards the hyperbolic may be counterbalanced by the desire to be believed.
· He suggests that this has less influence in cases where the narrator is not personally responsible.
(He doesn’t explain how he reaches the conclusion as to when such hyperbole is more or less likely. I’m not sure if this is based on prior research or simply his opinion.)
4. Errors of memory
· This presents even more risk according to the author:
· He states that if the witness regards the facts in a particular speculative or emotional light, their memory will be apt to adapt to this view.
· Details will get introduced or dropped out so as to “aid the harmonious effect” (
§6)
· He notes that even apart from special bias, the witness may fill in the pictures with wrong detail, or forget some aspects
(I agree with him on the high risk. I also agree that if emotion is involved that can influence – but as he notes, you don’t even need that to have wrong details accidentally incorporated into a memory, or other details forgotten. I wish he had gone into more of a discussion as to what factors let us have more confidence in the reliability of the memory, and what factors don’t. Also what they are based on. He discusses below the effects of time passing, but doesn’t discuss much about our capacity to accurate recall spontaneous events even minutes later. This has to be part of the analysis – though I don’t think it makes a difference in the results of this study, as per his conclusions below. It would be a factor going forward.)
He notes that we have to consider how these various sources of error may affect the evidence for a case of spontaneous telepathy.
He suggests four areas to examine:
(1) Aparticularstateoftheagent,
e.g.,thecrisisofdeath.
· He attributes low risk or error here in that it will usually be clear what happened to him. (
§8.)
(This is probably fair enough most of the time, will depend on the situation of course)
(2) Aparticularexperienceofthe percipient,
e.g.,theimpressionofseeingtheagentbeforehim invisibleform
· Often have nothing but his own word
· but for what is required, his word is often sufficient.
· The author is making the important distinction (that I have made often and been derided for incidentally) between:
o the perception of the experience, and
o the correct interpretation of the experience.
· He notes that the interpretation could be erroneous, such as which hallucination.
· He writes: “Theriskofmisrepresentationissmallestifhisdescriptionof hisexperience,oradistinctcourseofactionduetohis experience,has
precededhisknowledgeofwhathashappened totheagent.”
(I agree with him as to the difference between the perception and the interpretation. Should add onto this the error in memory problem: is: Is he remembering the experience correctly. Again, not crucial for the results of this study it seems but would impact future studies.)
(3) Thedateof(1).
· The author notes that when the description of the experience dates from a time subsequent to his knowledge of what has happened to the agent, there is the possibility that this knowledge may have made the experience seem more striking and distinctive than it really was. (
§9.)
· He notes they did not detect “definite” instances of this sort of inaccuracy:
(Will have to see what he means by this. I’m not sure if this is indicating that he is looking for evidence of these errors rather than evaluating the risk of these errors as the Cochrane research suggests is the more sound approach.)
· He writes: “Norwouldthefact(oftenexpresslystatedbythe witness)thattheexperiencedidnotatthetimeofits occurrencesuggesttheagent,byanymeansdestroy—thoughit wouldofcourseweaken—thepresumptionthatitwastelepathic”
(Not quite sure what he’s getting at: Is he referring to times where the person reports the general experience before learning of the agent but doesn’t make the association with the agent until after?)
· His study has adopted an arbitrary 12 hour difference between the experience of the agent and the percipient (and notes that most occur much closer in time) –
· He will reject any case where the percipient’s experience preceded some grave event occurring to the agent, if at the time of the percipient’s experience the state of the agent was normal:
(seems to be saying not looking at pre-cognition?)
(4) Thedateof(2)
· The author attributes the biggest risk of misstatement to the matter of dates: That after the fact the instinct to simplify and be dramatic can lead to wrongly assuming the dates matched.
(while I agree there is risk here, is it bigger than re: the details?)
· He notes that often the date of the event re: the agent is published, and is often close enough to the percipient’s experience for the date to be safely recalled.
(again, not clear upon what he is basing this. Is it just common sense or the results of research?)
· Notes that the more time that passes the greater the risk of assuming the experience fell on the critical day. Certainty increases with time not diminishes.
· If the coincidence was then and there noted (does he mean documented or simply pointed out.) and if attention to others was called to it, may be possible to present strong enough case for its reality, even after lapse of considerable time.
He suggests that these evidential conditions can be put on a graduated scale: (
§14)
· Second hand evidence excluded with the exception of when witness well acquainted with the original witness
· in “transmitted evidence” (not sure what this refers to – probably clear in later chapters) risk of error very high (suggests he is looking at risk rather than actually spotting evidence)
o frequent error is the shortening of the chain of transmission: 2nd or 3rd hand info being represented as first-hand – alleged coincidence suspiciously exact
· Study attempts to separate cases according to evidential value and focusing on cases with strongest prima facie probability that the essential facts are correctly stated.
· Even when facts correctly reported, strength vis-à-vis telepathy differ according to class:
o purely emotional impressions and dreams are very weak.
· doesn’t go into the other categories here (
§16)
· Also depends on the mental qualities and training of the witnesses:
o doesn’t say what these are or how it affects things
(Given that he says that previous studies hadn’t looked at these things, it would have been interesting if part of this study was to document how the various risks that he identifies affects the reports - that is, if this study’s purpose was more to figure out what factors affect reports and how. as it stands, it is difficult to determine, based on what he’s writing here, upon what he is basing his conclusions here.)
· Each case evaluated on its own merits, by reference to variety of points.
· Number of coincidences:
o considers this all-important
§ few might be accidental
§ impossible to apply that hypothesis throughout
§ can’t just sweep evidence under the rug by mere general appeal to untrustworthiness of human testimony
(He’s essentially correct in his approach in terms of evaluating risk of bias vs. just blindly calling all human testimony untrustworthy. He’s gone into a number of factors related to how we should evaluate reliability though there is much more that could be said here. That said, for the time, he’s done a pretty good job of identifying areas of concern.)
§ If it is to be explained away, must be met in detail
(This line is potentially problematic but he doesn’t make the error I feared here in terms of his conclusions – I suspect this is a line that many people since have misinterpreted- reading into what he wrote something that he didn’t mean: ie: it appears to suggest that you have to identify the alternate cause to rule out telepathy, but I don’t think, given his conclusions below, that he really does mean that.)
§ “andthisnecessitatesthe confrontingofthesinglecause,telepathy,(whose
àpriori improbabilityisfullyadmitted,)withamultitudeofcauses, moreorlessimprobable,andincumulationincredible” (
§17)
(He doesn’t elaborate here on the number of coincidences part, I assume there is more elsewhere so I can’t comment on this now.)
§ “
§18.Withalltheirdifferences,thecasesrecordedbear strongsignsofbelongingtoatruenaturalgroup;andtheir harmony,alikeinwhattheydoandinwhattheydonotpresent, isveryunlikelytobetheaccidentalresultofamultitudeof disconnectedmistakes”
(this is a conclusion – will have to see how he gets there- also have to watch out for whether this is post-hoc exploratory grouping – Is he grouping these cases according to the categories he’s set out above, or are these post-hoc categories. For that matter, were those categories he laid out above decided before the study or after? As I said above, I hope part of this study documents how the factors he related above result in different types of accounts.)
Author’s Conclusions on the evidentiary value of the study
§ The author gives a preview of his conclusions.
§19.Butthoughsomemayregardthecumulativeargument hereputforwardforspontaneoustelepathyasamountingtoa proof,theproofisnotbyanymeansofanéclatantsort:muchoftheevidencefallsfarshortoftheidealstandard.Still,enough hasperhapsbeendonetojustifyourundertaking,andto broadenthebasisoffutureinquiry
§ The author therefore concludes:
o much of the evidence doesn’t meet the highest quality standards
o the evidence provided is strong enough, however, to indicate that further – higher quality – studies be done.
(This is the same kind of conclusions we see all over the place in parapsychology – there is broad agreement (including myself) that the higher risk of bias studies present the justification for pushing forward with lower risk of bias cases.
The fact the author acknowledges that most of the evidence does not reach the low risk of error status does not mean that his research is “flawed” or “sloppy”. As he stated earlier, this area of research is very difficult to capture in low risk of error ways.
I agree with the author that the value of the study is that it justifies further studies.
The author, sees promise in future studies, but does not promote this study as being sufficient in and of itself. Rather it is suggestive and warrants further research. The author does not consider it strong evidence – which puts the author at odds with some forum comments on this study which seem to consider it to be strong evidence. The author is not trying to make this study out to be more than it is, neither should readers.
§ As the author writes: this study is “atleastaninstallmentof whatisrequired.” (
§20)
(the question then begs, since 1886 what evidence has been presented that improves on this. One poster in the forum suggested that this study remains to this day the best of the field on this topic. So we still await the higher quality studies that the author suggests should follow.)
Arouet’s General Conclusion on Chapter IV
Overall, especially given when it was written, the author has identified a number of areas that lead to a risk of bias. It is curious that when skeptics on this forum raise similar issues, using similar words, they are criticized for it.
The study of these biases has advanced since that time, and some ideas that he raises have become more fleshed out over time, but presuming his [Edit: I cut off there, presumably because I at that point realised that I had just wasted a huge amount of time analysing a synopsis!]