[extropy-chat] Re: Damien grants psi evidence

[Brett Paatsch]

Damien Broderick wrote:

> At 09:55 AM 12/18/2004 +1100, Brett wrote:
>
>>I agree with Bill's comments on "cold reading" ... The judge
>>should be scoring not just matches as the judge sees them but actual
>>matches that are objectively the same matches that the viewer makes.
>>
>>Take the judges subjectivity about what is a match out of the equation
>
> This makes zero sense to me. ...

Perhaps in my above comments I was being too lazy.

As I was reading Hyman's report I excerpted the following into
a side file, I suspect there a good chance that Bill read it too:

"So, both the viewers and the judge quickly became convinced of
the reality of remote viewing on the basis of the uncanny matches
between the verbal descriptions and the actual target sites. The
experimenters received a rude awakening when they discovered that,
despite the striking matches observed between target and verbal
description, the judge had matched the verbal protocols to the wrong
target sites. When all parties were given the results the subjects could
not understand how the judge could have matched any but the actual
target site to their descriptions. For them the match was so obvious
that it would be impossible for the judge to have missed it. The judge,
on the other hand, could not accept that any but the matches he made
could be paired with the actual target sites.

This phenomenon of subjective validation is pervasive, compelling
and powerful. Psychologists have demonstrated it in a variety of
settings. I have demonstrated it and written about in the context of
the psychic reading. In the present context, subjective validation
comes about when a person evaluates the similarity between a
relatively rich verbal description and an actual target or situation.
Inevitably, many matches will be found. Once the verbal description
has been judged to be a good match to a given target, the description
gets locked in and it becomes virtually impossible for the judge to see
the description as fitting any but the original target.

I can imagine that the preceding paragraph might strike a reader as
being unreasonable. Even allowing for subjective validation, the
possibility that  a viewer might accurately come up with secret code
words and a detailed description of particular gantry is quite remote
on the basis of common sense and sophisticated guessing. I understand
the complaint and I realize the reluctance to dismiss such evidence out
of hand. However, I have had experience with similarly compelling
prima facie evidence for more than a chance match between a
description and a target. In the cases I have in mind, however, the
double blind controls were used to pair descriptions with the true as
well as with the wrong target sites. In all these test cases with which
I am familiar, the unwitting subjects found the matches between their
descriptions and the presumed target equally compelling regardless of
whether the presumed target was the actual or the wrong one.

What this says about operational effectiveness, is that, for evaluation
purposes, half of the time the viewers and the judges should be
mislead about the what was the actual target. In these cases, both the
interrogator and the viewer, as well as the judge, have to be blind to
the actual targets. Under such conditions, if the judges and the others
find the matches between the verbal descriptions and the actual targets
consistently better than the matches between the verbal descriptions
and the decoy targets, then this would constitute some evidence for
the effectiveness of remote viewing. I can confidently predict, regardless
of the outcome of such an evaluation, that many of the verbal
descriptions when matched with decoy targets will be judged to be
uncanny matches."

Damien again, but with a protocol I'm not certain is exactly what
Utts used:

> A randomizer machine draws out four possible targets without
> telling anyone which they are: a fluffy brown dog playing near a grassy
> knoll; a gorgeous photo of the Millau Bridge
> (http://news.bbc.co.uk/2/hi/in_pictures/4095037.stm ), a plate of
> bacon and eggs, and a yacht surging on the high seas. One of these
> becomes the actual test target; nobody yet knows which. The RVer
> does his or her stuff, filling up pages of subjective responses. The
> judges look at the viewer's report: several drawings of a witch's
> hat, of a church steeple, of a man leaning over a rail vomiting. None
> of these matches any of the pictures.

Not exactly, no. Not as exactly as would have been possible had
the RVer instead been shown copies of the four pictures and asked
which of the four best matched the one at location X that they could
not see directly.

> The test is obviously a failure.

Or too demanding a test, too high a hurdle, for what might be a real
but subtle and context sensitive phenomena.

> But wait, here's where the judges come in.

Either to make their judgements separately, or collectively.

> They know that the
> mind elaborates partial information by `best guesses', which are
> somewhat mutable; perception is always a construct; we home
> in on a piece of trash  that looks like a scurrying cat in twilight.

I grant that they'd know that. Certainly most folks with some
psychology training or understanding would.

> So the judges muse. Hmm, the puking guy might have eaten the
> eggs, or might associate dogs with dog shit. The bridge, from some
> direction, could look  like a witch's hat. But then both the
> hat and the steeple structurally resemble the view of the yacht
> more than they do the bridge piers, and the barfer is perhaps
> an association to a rough sea trip. So the judges rank the options 1.
> yacht, 2. bridge, 3. eggs, 4. dog. Their majority
> vote selects yacht.

But do the judges do their musing individually and then vote
without discussion as to the basis of their choices or are they allowed
to converse, and perhaps explain and/or offer justifications for their
choices or preferences in rankings such that they might influence
each other?  

Judgements made in social contexts can differ from those made
individually. Why, or if this could matter in this case I don't know,
but it might.  

> They go on their way, and are never heard from again.

Sure, thus avoiding any interaction with the RVer at all.

> Perhaps the RVer is also shown the possible targets, and selects one as
> closest to the blurry psi perception.

Ok. But as the RVer would either be shown or not shown,
by perhapsing like that we have two classes of protocol to
consider instead of one.

>  The target is disclosed to the experimenter who has also been blinded,
> and to the RVer. (Or sometimes *only* to the experimenter.)
>
> Gee, it's the yacht. We'll take that to be a hit.

If the RVer has selected the yacht from the four possibilities and the
target is a yacht then the judges rankings don't matter as there is nothing
left for them to to judge. The experimenter alone could see that the target
and the RVers selection are the either the same or not the same.

As I understand it the judges are only there to increase the sensitivity of
detection enough to pick up weak effects (nothing inherently wrong with
that insofar as it goes, just so long as that is all the judges are really 
doing and they are not introducing a confounding variable of another
sort. When the psi effect itself is posited to be weak but real then its
all the more important that other confounding effects including weak 
ones are not unnecessarily introduced. Perhaps people - both judges
and RVers musing or free associating are more likely to see some 
things, relationships, patterns etc than others and so our musings are
not truly random but biased and perhaps biased similarly by our
common evolutionary or cultural heritage even when we are not 
overtly aware of any bias).

In the excerpt from Hyman above, I got the impression that the judge(s)
might be introducing other effects.  I'm not completely convinced that
Hyman's criticism in relation to the judges specifically applied to Utts
particular findings, its rather just that I can't remember her denying it
in her reply to him so I creditted it as more likely to be valid.

Alas, (or "sigh" :-), human judgements are made in such flimsy ways.
That is why the burden of proving unusual stuff is so heavy. One has
not only got to persuade, one has to hold the interest and the
attention.

> How does fishing come in? What cold reading? Of whom?

It is not really fishing or cold reading but more what Hyman above calls
subjective validation. There is in people a bias to finding something
rather than nothing. To succeed rather than fail. To being included
rather than excluded. To being significant rather than insignificant.
This bias can often make us over estimate agreements and underestimate
disagreements. Hymans point seemed to me to be that the agreements
on matching might not actually be agreements on the same match. And
he offered a suggestion (last para quoted above) for eliminating the
problem. 


Its possible that groups of judges might converge on a similar attractive
or plausible sounding stories for ranking things a particular way even if
there is no real rational reason for prefering that story or ranking over
other plausible ranking alternatives that might have been offered.

I think cold reading arises to some extent because people want to be
included, significant, agreeable, more than the alternatives, but "cold
reading" isn't a base level explanation of anything its just an instance
of a wider class of observed phenomena where what Hyman calls
subjective validation seems to come into play. Whenever there is
inherent ambiguity more people will interpret the data in ways that
are positive or elevating to them and their world view (and their
relations with others) than the alternatives.

Brett Paatsch