In Sri Lanka, a toddler one day overheard her mother mentioning the name of an obscure town (“Kataragama”) that the girl had never been to. The girl informed the mother that she drowned there when her “dumb” (mentally challenged) brother pushed her in the river, that she had a bald father named “Herath” who sold flowers in a market near the Buddhist stupa, that she lived in a house that had a glass window in the roof (a skylight), dogs in the backyard that were tied up and fed meat, that the house was next door to a big Hindu temple, outside of which people smashed coconuts on the ground. Stevenson was able to confirm that there was, indeed, a flower vendor in Kataragama who ran a stall near the Buddhist stupa whose two-year-old daughter had drowned in the river while the girl played with her mentally challenged brother. The man lived in a house where the neighbors threw meat to dogs tied up in their backyard, and it was adjacent to the main temple where devotees practiced a religious ritual of smashing coconuts on the ground. The little girl did get a few items wrong, however. For instance, the dead girl’s dad wasn’t bald (but her grandfather and uncle were) and his name wasn’t “Herath”—that was the name, rather, of the dead girl’s cousin. Otherwise, 27 of the 30 idiosyncratic, verifiable statements she made panned out. — Wayfarer
Just to prove to you that my rejection of these reports 'suggesting' that remembrance from past lives is not based on blind prejudice, or scientific orthodoxy, or whatever irrelevant standard you're trying to portray me as holding, let's go through the above report.
In Sri Lanka, a toddler one day overheard her mother mentioning the name of an obscure town (“Kataragama”) that the girl had never been to.
Great, already that the memory began with an elicitation from an external source. Moreover, it was part of a conversation which was recorded afterwards. This immediately means you can't distinguish conversational priming effects from those which arise from remembrance of past lives.
The girl informed the mother that she drowned there when her “dumb” (mentally challenged) brother pushed her in the river, that she had a bald father named “Herath” who sold flowers in a market near the Buddhist stupa, that she lived in a house that had a glass window in the roof (a skylight), dogs in the backyard that were tied up and fed meat, that the house was next door to a big Hindu temple, outside of which people smashed coconuts on the ground.
Did
the girl believe that her brother was 'mentally challenged' or that he was just 'stupid'? The '(mentally challenged)' bit is obviously some post processing of the report, we don't know if the girl actually remembered the brother as being an idiot or being mentally challenged or disabled in some non-specified way. This lack of specification allows the collator of the reports to substitute in 'mentally challenged' for 'dumb' in a just-so story. More formally, this is spending many 'researcher degrees of freedom' to tailor the post-processed (not original!) account to other things. On your standard, you would interpret low IQ, clumsiness, dyspraxia, any developmental disability effecting the mind, as equally being vindicated by the child saying she had a 'dumb' family member. But these details really matter. They matter because someone who had these
specific memories with
specified relationships in them would be having a typical event of memory; a young girl will probably know in what sense someone is 'dumb', but all we have here is that dumb was mapped to 'mentally challenged' during the post-processing of the report, with a vague status on precisely how much elicitation the mother treated the daughter too. You can only 'weasel out' of this with the stipulation that past life memories are qualitatively different from memories... IE they are not actually memories in the usual sense of the term.
This matters. A lot. You can give 'psychic readings' to people where you
say Barnum statements and
elicit specified responses, often people will
remember you saying the
elicited response rather than the Barnum statement. The report will say that 'The psychic predicted so many specific things about me!' and the person given the reading will have that memory, but the causal mechanism was one of elicitation rather than memory in a vacuum.
Herath is a unisex name, given that we do not know the degree of elicitation here, we cannot associate 'Herath' with 'bald father' accurately (also 'bald father', in a town with a Buddhist history, how specific!). We
should have the same response to 'bald father called Herath' to 'mother with long brown hair called Herath', and it does not matter for the truthiness of the story which is which. But it absolutely does matter for the purposes of ascribing cause; the two are exchangeable for the purposes of the narrative validity, but are not exchangeable in terms of connection to a memory. You'd be saying the same thing for both, but they're different reports which would have been fit to the facts differently.
A marketplace near a stupa makes a lot of sense, they're going to be central features of a town.
For instance, the dead girl’s dad wasn’t bald (but her grandfather and uncle were)
Yes, surely the memory averaged over the contents of male relatives, rather than being rather specific
upon elicitation like usual memories. This is a retrojection of memory validity which just wasn't there, and has no causal mechanism associated with it. If you could tell me
why it makes sense for 'ancestral memories' to sometimes average over male heritage lines and sometimes not I'd love to know.
The same goes for having dogs in the backyard being fed meat, this is common. The same goes for coconuts on the ground in marketplaces...
Anyway, the minimal criterion for one thing X being an indicator of another Y is that X is more probable (or less probable) given Y. One needs to
assume that the relationship between the events in the town and the events the girl came up with are one of memory in order to
assign that one is more probable given the hypothesis of memory. This goes in completely the wrong direction, what you should do when reviewing such reports is to ask the question "How likely is the scenario envisaged given how likely the scenario is to happen?'.
And when we ask 'how likely is the scenario to happen', the scenario needs to be highly specified, and not fit to the facts on an interpretive basis (which is what was done during the 'search' for the events of the memory).
Moreover, you also need to look over all possible remembrance events, if most are filtered out due to evidential standards, but some are left in without a pre-specified standard of validity (not which 'seem hardest to explain), you still have the problem of why are there so few reports which meet a higher evidential standard. Is the mechanism for generating the reports with a lower evidential standard substantially different for the mechanism generating the reports with a higher evidential standard? Or are they
post selected for demonstrative purposes (which is what is being done here!)
If I took all the obese people in the world and looked at their chance of heart disease, I would vastly over-estimate the chance of heart disease in the general population if I did not take into account the fact that I have filtered the population. The same thing happens here, the reports of higher evidential standard say nothing about those which have a typical evidential standard; IE, most reports of such things are so easy to recognise as incredibly flawed that they are immediately removed from the study. To put it in plain words, the fact of the matter is that even if you grant that the memories considered in the higher evidential standard group are more likely to be remembrance of past lives than the ones in the lower evidential standard group, most reports of past lives are
still too flawed to use as evidence.
And what is the standard for the ones in the higher evidential standard group? Not just the quality of elicitation, surely, one also wishes to check if they are true. Now, when they are true, we have a tiny subset of reports which purportedly describe real events. We now need to ask the question: does the probability of the real event increase when the event is described or elicited? But of course, since this pre-selection by coincidence of report with real events has happened, one can never ask this question of the data. The very criterion with which you would establish a
mechanism of past life remembrance is excluded through the filtration of the study to the ones with a 'higher evidential standard'; which apparently is literally just the events elicited/described by the remember happened somewhere at some time under some interpretation (researcher degrees of freedom).
You might say that they happened in place X in way Y, but one needs to aggregate over the reports to get the real picture of what is going on. The relevance here is that things which happened in place X in way Y and were described as happening in some way related to X and some way related to Y is not the same criterion of validity as events X happening in way Y
in the report (also note the priming/elicitation memory confusion here, it interacts!). The 'in some way' matters, as this is an inherent part of the filtration procedure; it has so many 'researcher degrees of freedom' that some fit is bound to happen for some reports. It is a mismatch inherent in the selection procedure for validity, you spend all the information you have on establishing the coincidence of real event with described event, not the conditional report validity (which is the minimal criterion for informativeness of X on Y... which is required for X to be memories of Y)
You might say I'm being too cautious, I would not hold up my beliefs in every-day life to this standard. And you're right, but the every-day is every-day, we
need to have higher standards of evidential validity when considering questions as big as 'can we remember past lives?'.