Improbability Principle

For what its worth, I belive it was Rupert Sheldrake who pointed out, that after ww2 ended and home milk deliveries resumed. The Great Tits resumed thier early morning raids. The problem being, all the prewar Great Tits had died of old age or other causes.
(The war lasted longer than the life span of a Great Tit.)

For what its worth, I belive it was Rupert Sheldrake who pointed out, that after ww2 ended and home milk deliveries resumed. The Great Tits resumed thier early morning raids. The problem being, all the prewar Great Tits had died of old age or other causes.
(The war lasted longer than the life span of a Great Tit.)

Hope I'm not breaking any copyright rules but I thought it was worth posting a section of a much longer essay by Sheldrake available on his website.

There are other examples of the spontaneous spread of new habits in animals and birds which provide at least circumstantial evidence for the theory of morphic resonance. The best documented of these is the behavior of bluetits, a rather small bird with a blue head, that is common throughout Britain. Fresh milk is still delivered to the door each morning in Britain. Until about the 1950s, the caps on the milk bottles were made of cardboard. In 1921 in Southampton, a strange phenomenon was observed. When people came out in the morning to get their milk bottles, they found little shreds of cardboard all around the bottom of the bottle, and the cream from the top of the bottle had disappeared. Close observation revealed that this was being done by bluetits, who sat on top of the bottle, pulled off the cardboard with their beaks, and then drank the cream. Several tragic cases were found in which bluetits were discovered drowned head first in the milk!

This incident caused considerable interest; then the event turned up somewhere else in Britain, about 50 miles away, and then somewhere about 100 miles away. Whenever the bluetit phenomenon turned up, it started spreading locally, presumably by imitation. However, bluetits are very home-loving creatures, and they don't normally travel more than four or five miles. Therefore, the dissemination of the behavior over large distances could only be accounted for in terms of an independent discovery of the habit. The bluetit habit was mapped throughout Britain until 1947, by which time it had become more or less universal. The people who did the study came to the conclusion that it must have been "invented" independently at least 50 times. Moreover, the rate of spread of the habit accelerated as time went on. In other parts of Europe where milk bottles are delivered to doorsteps, such as Scandinavia and Holland, the habit also cropped up during the 1930s and spread in a similar manner. Here is an example of a pattern of behavior which was spread in a way which seemed to speed up with time, and which might provide an example of morphic resonance.

But there is still stronger evidence for morphic resonance. Because of the German occupation of Holland, milk delivery ceased during 1939-40. Milk deliveries did not resume until 1948. Since bluetits usually live only two to three years, there probably were no bluetits alive in 1948 who had been alive when milk was last delivered. Yet when milk deliveries resumed in 1948, the opening of milk bottles by bluetits sprang up rapidly in quite separate places in Holland and spread extremely rapidly until, within a year or two, it was once again universal. The behavior spread much more rapidly and cropped up independently much more frequently the second time round than the first time. This example demonstrates the evolutionary spread of a new habit which is probably not genetic but rather depends on a kind of collective memory due to morphic resonance.

Does anyone have a primary source for the claim about blue tits' behaviour starting in 1921? Sheldrake's essay contains no references at all.

I quickly checked the British Newspaper Archive to see what came up and the earliest mention of it I can find is 1943 (by the way, milk deliveries did not stop in Britain during World War II).

Does anyone have a primary source for the claim about blue tits' behaviour starting in 1921? Sheldrake's essay contains no references at all.

I quickly checked the British Newspaper Archive to see what came up and the earliest mention of it I can find is 1943 (by the way, milk deliveries did not stop in Britain during World War II).

It's mentioned in this published paper from 1949, though from a quick look it's not obvious what the authors' source was:
https://www.britishbirds.co.uk/wp-conten...7_A059.pdf

It's mentioned in this published paper from 1949, though from a quick look it's not obvious what the authors' source was:
https://www.britishbirds.co.uk/wp-conten...7_A059.pdf

Thanks, that's great  No mention of the situation in Holland, though. I'll try emailing Sheldrake for a source on that claim

Actually, I suppose the 1921 example just represents the earliest sighting mentioned in response to the questionnaire circulated by the authors of that paper.

I see there is an addendum at the end saying the authors had been informed that the phenomenon wasn't known in Holland, "where the use of milk bottles with metal foil stoppers is widespread".

Google Books certainly doesn't make it easy to check references, but the information about Holland comes from a sequel to that paper, published by the same authors in 1951:
https://britishbirds.co.uk/wp-content/up...6_A069.pdf

I've been thinking about ways to counteract our tendency to treat examples selected post hoc as somehow representative and to intuitively guess at probability/frequency, rather than measure it.

One idea I've mentioned before goes back to the idea of falsification/disconfirmation - that is, ask "what do I expect to see in the absence of the effect?" However, this runs into the problem I brought up in the OP, where people guess at what they expect to see based on intuitively estimating the probability of the specific event. It has been discovered that if you measure the frequency of an extraordinary event in the absence of an effect instead, the frequencies are much, much higher than our intuitive estimations. So I'm thinking that the next bit of advice is to measure that frequency rather than guessing at it. This made a huge difference in medicine when it turned out that guesses about what would happen without treatment or simply giving current treatments more time, were wildly inaccurate. Once you measure what happens by adding control groups, the frequency of what happens in the "treatment" group often turns out not to be remarkable or different. 

And speaking of frequency, it probably also wouldn't hurt to measure the frequency of extraordinary events prospectively, rather than depending upon our usual method of examples selected post hoc, especially in the form of anecdotes (an anecdote is an undocumented story, and something is undocumented if it is based on recollection rather than a more permanent record of the events made prior to any feedback). 

The NDE research has been a good example of this disconnect. NDEs came to our attention because a (relatively) tiny set of stories were regarded as extraordinary. And people went about asking for these same kinds of stories to form a collection (e.g. IANDS), which produces a highly selected sample. When the samples are collected prospectively instead, the frequency of extraordinary experiences which are similar to the stories on IANDS is very low relative to all the experiences people have. And the stories selected post hoc turn out to be unrepresentative of the population of experiences. That is, the process of asking for "NDEs" selects for a set of experiences which are memorable and have an element which can be regarded as spiritual. This is taken as meaningful when people start to assume that most experiences had by people during medical crises are of this type, and that the experiences which are not of this type are different in quality (i.e. they aren't life-altering, realer than real, richly detailed, veridical, emotional, etc.). Prospective studies instead show that experiences which have an element which can be regarded as spiritual don't differ from the other experiences in terms of whether they are memorable, life altering, realer than real, richly detailed, veridical, emotional, etc. 

Ask what you'd expect to see in the absence of the proposed effect.
Measure the frequency of those expectations.
Collect representative samples of extraordinary experiences.

Other ideas with respect to addressing the issues raised in the OP?

Linda

I've been thinking about ways to counteract our tendency to treat examples selected post hoc as somehow representative and to intuitively guess at probability/frequency, rather than measure it.

The NDE research has been a good example of this disconnect. NDEs came to our attention because a (relatively) tiny set of stories were regarded as extraordinary. And people went about asking for these same kinds of stories to form a collection (e.g. IANDS), which produces a highly selected sample. When the samples are collected prospectively instead, the frequency of extraordinary experiences which are similar to the stories on IANDS is very low relative to all the experiences people have. And the stories selected post hoc turn out to be unrepresentative of the population of experiences. That is, the process of asking for "NDEs" selects for a set of experiences which are memorable and have an element which can be regarded as spiritual. This is taken as meaningful when people start to assume that most experiences had by people during medical crises are of this type, and that the experiences which are not of this type are different in quality (i.e. they aren't life-altering, realer than real, richly detailed, veridical, emotional, etc.). Prospective studies instead show that experiences which have an element which can be regarded as spiritual don't differ from the other experiences in terms of whether they are memorable, life altering, realer than real, richly detailed, veridical, emotional, etc. 

Other ideas with respect to addressing the issues raised in the OP?

Linda

Linda,

I would have liked your post -except for the obfuscation exhibited in the above paragraph.

In terms of reported human behavior, within all cultures, you imply that NDE's and paranormal experiences started in a recent time-frame and are reported by only a few people.  When in fact; the elephant in the room is that paranormal events are reported in all places and times of recorded history.  The study of stone and bronze age cultures are replete with reports.

So before measuring something (merelogy) you need to define some stuff.  First if you say - NDE/paranormal is a category, then the category needs a specific definition that will lead to accurate measurement!  So let's look at a RCA (root cause analysis).

Paranormal events are events where information transfer (both formal information or semantic information) occurs without a signal and a physical channel where the signal can forwarded to a receiver capable of decoding the message.

Dreams, intuitive senses, instinct, understanding from imagination, understanding from empathy of abstract circumstances - all qualify.

Ok now you can start working on frequency with an unrestricted environment - all places and all time - and an informational definition of what sorts  - normal perception - from paranormal perception.  Take a sample of people who have taken meaning that has changed probable actions in their lives; and then project it on the environment; resulting get a frequency of occurrence.  It will be in the hundreds of millions.

Note that spirtual (what ever the frack that is in a formal defintion) IS NOT a variable in this methodology - just a clear measurable definition of mutual information being generated with a physical signal.