Re: Cambrian Explosion

Kevin O'Brien (
Sun, 7 Mar 1999 09:26:36 -0700

>>>I thought I would throw my two cents on this, hope you
>>>don't mind :).
>>As a worthy opponent you are most heartily welcome. Come, let us toast
>>each other's bravery and fortitude. Slantha!
>This sounds like something one hears just before being skewered ;-).

Perhaps. My Celtic ancestors, when they met on opposite sides of a battle,
nonetheless complimented and toasted one another's bravery just before they
got down to the business of slaughtering one another.

>>>This insistence by some that absence of
>>>evidence is not evidence of absence has always bugged me
>>>since, for one thing, it seems totally contrary to the
>>>way science normally works.
>>Not really. Science is based on the idea that you need definitive,
>>unambiguous evidence that specifically validates or refutes hypotheses.
>>I explained to Mark in my latest reply, you can expect one of three basic
>>results from any experient: positive evidence that verifies the
>>positive evidence that refutes the hypothesis, or no positive evidence,
>>which is then called negative evidence. It is negative evidence that the
>>appeal from ignorance specifies when it refers to "absence of evidence".
>>Practically speaking, negative evidence is either no evidence whatsoever
>>(such as from a failed experiment) or evidence too ambiguous to interpret;
>>it is not positive evidence that refutes the hypothesis (hence producing
>>an "absence" of positive evidence that would have verified it instead).
>I'm afraid I'm getting confused. Perhaps I'm slow :). It always
>helps me to consider examples, preferably different from those
>under dispute :)
>When I think of lack of evidence in a scientific setting
>I'm thinking mainly of two possibilities:
>1) There is a particular piece of evidence which would really
>bolster a theory were it available. Unfortunately, it is not.

>2) The theory makes a specific prediction that such and such
>evidence should be found in such and such a fashion. But this
>evidence is not found.
>How would you define these two "lack of evidence" situations
>using your terminology?

Both are clearly cases of negative evidence, since we lack positive evidence
that would support the theory, but we do not have any evidence that refutes
the theory either. A failed prediction can caste doubt on a theory, but as
long as the result obtained does not actually contradict the theory, in and
of itself it does not constitute positive evidence that refutes the theory.
In alot of cases, however, the result does tend to contradict the theory in
some way, thus constituting positive evidence that refutes the theory.

>It might help to think of two concrete examples for each.
>1) Newton's first law says a particle moving in a straight
>line at constant speed will continue to do so *forever* unless
>acted on by a force. Direct evidence for this would be the
>observation of a particle moving in a straight line at
>constant speed in absence of forces. I believe evidence of
>this type is lacking. Since the lack of evidence arises because
>the critical experiment cannot be performed we should, I think,
>be neutral regarding this particular lack.

I'm not sure what you are saying here, but it sounds as if you are saying
that since the definitive experiment cannot be done, we cannot say one way
or the other whether the first law is really true. In point of fact,
though, spacecraft have largely verified the first law, and even before them
there were simple experiments that any school kid could do that did not
contradict the first law. So, unless there are more sophisticated
experiments that have contradicted the first law, I'm not sure we need to
perform the definitive experiment.

>2) For an example of the second type we'll propose an indirect
>test of the law of inertia and a direct test of the inverse
>square law. Specifically we propose to measure how far the
>moon "falls" in some time interval. By "fall" I mean more
>precisely the distance that the moon deviates from its natural
>(inertial) motion. The experimental phase of our test was
>accomplished by astronomers before Newton published his universal
>law. Thus, all Newton had to do was see if he could predict the
>results. Turns out he failed to do so, resulting in a five year
>delay in publication. Apparently Newton thought that the absence
>of this evidence was a critical blow to his theory. Later someone
>(not Newton) repeated the measurements finding the initial results
>to be in error. When Newton obtained the revised calculations he
>found they matched his predictions beautifully. He published.

So that really wasn't positive evidence that refuted his theory, it was
negative evidence based on bad calculations (the equivolent of a failed
experiment). Unless there is more to the story than you said, it sounds
more like a case of failed nerves. It seems to me that rather than seeing
it as clear positive evidence that refuted his theory, Newton allowed that
negative evidence from that failed experiment to shake his confidence. Once
he discovered the source of his error and obtained the correct results, he
had positive evidence that supported his theory and his confidence returned.
Had he instead retained his confidence and tried recalculating the figures
he might have gotten the right results sooner. (From what I've read about
Newton, he was prone to self-doubt.)

Besides, as I shall expand upon later, Newton was in the development phase
of his theory, when in fact an absence of evidence can refute a theory, not
because it constitutes positive evidence that refutes the developing theory,
but because in the absence of positive evidence that supports it, a
developing theory is little better than speculation.

>>>If someone proposes a theory
>>>it is natural to ask that person what the evidence is for
>>>that theory. But if absence of evidence is not evidence of
>>>absence then what, pray tell, is the purpose of such a
>>To test the theory against physical reality, you silly person ;-).
>Exactly. And if the person replies that evidence is lacking, how
>well does his theory score on this test?

My point was that there is a difference between the evidence that is used to
develop the theory and that which is used to test the theory. If we are
discussing a legitimate theory, then the evidence that was used to develop
it can automatically be considered positive evidence that verifies the
theory. If there is no such evidence then we are not talking about a theory
but speculation, in which case the question of absence of evidence doesn't
even apply. However, the lack of evidence used to test the theory may
simply be the result of not having tested the theory yet. It may even be
technologically impossible to test the theory at this time. As such, this
lack of evidence cannot tell us anything about whether the theory is viable
or not, and we must instead rely solely on the evidence that was used to
develop the theory, which I have already stated can be considered positive
evidence that verifies the theory.

>>any proper theory must have some supporting evidence from the start
>>(otherwise it is only speculation), this is not what verifies or refutes
>I fear we may be talking past each other as I find the above to be
>an affirmation of my point :). If a proper theory must have some
>supporting evidence then clearly the absence of same would be detrimental
>to that theory.

In a manner of speaking; it wouldn't be a theory at all. But in this case
you are in a situation where you either have positive evidence that verifies
the theory or you don't have a theory at all. The options of negative
evidence or postive evidence that refutes the theory are not possible in
this case, so the absence of evidence question cannot even be applied.

>Once we establish this then the rest seems a matter
>of degrees. Some theories have more evidence then others. These are
>more firmly established. Why?

I view theory verification as more of a sigmoidal curve. The initial period
of slow verification is the peiod when the theory is being developed. Once
enough evidence is collected to convince other researchers that the theory
is real, verification rapidly increases as the theory is tested. Finally,
when enough evidence has been collected to virtually prove the theory beyond
a reasonable doubt, verification drops off rapidly, but never really stops.
In any event, the inflection point between developing a theory and testing a
theory can be a bit fuzzy, but the two processes are distinct enough that
one can readily recognize when a theory is in one phase or the other. My
point is that when a theory is in the development stage, it takes very
little to refute it; even the absence of evidence can refute it, since at
that stage you must have evidence to even be a theory. During the testing
phase, however, things are not that simple. The theory now has enough basic
evidence to support it that a lack of (new) evidence no longer jeopardizes
it. What is needed now to prove it false is positive evidence that refutes
it. As long as the basic supportive evidence that was used to develop it
remains intact, the lack of any new supportive evidence will not refute it.
What is needed is evidence that refutes the supporting evidence already in

>>Instead you use the theory to make predictions about experimental results,
>>then do the experiments to try to get those results. As such, only
>>evidence that directly and unambiguously verifies or refutes the
>>will have any impact on the theory. Negative evidence, which by
>>cannot either verify or refute the prediction, can tell you nothing about
>>the theory. As such, any attempt to use this "absence of evidence" to
>>refute the theory is a fallacy.
>>Creationist "theories" tend to be speculation disguised as theory, so they
>>can be refuted without testing them by revealing their lack of supporting
>>evidence. This doesn't really involve the appeal from ignorance fallacy,
>>however, because the speculations-disguised-as-theories don't really
>>>Suppose that: (a) there is an absence of evidence
>>>that I robbed Fred's Bank on High Street yesterday....
>>In other words there is no positive evidence that proves OR disproves that
>>you robbed that bank.
>No. IOW there is no evidence I robbed the bank. My fingerprints
>weren't found, no eyewitness placed me in the bank. etc.

Since the absence of evidence is defined as a lack of positive evidence in
any form, your statement that "there is an absence of evidence that I robbed
Fred's Bank on High Street yesterday" means that there is no positive
evidence that proves that you robbed the bank AND there is also no positive
evidence that proves that you didn't rob the bank. As any forensics expert
can tell you, the lack of positive evidence that you robbed the bank (no
fingerprints, no eyewitnesses, etc.) does not constitute positive evidence
that you did not rob the bank (you could have worn gloves, you could have
worn a disguise, etc.). In some judicial systems you could still be
convicted if you cannot provide positive evidence that you could not have
robbed the bank (you were in Paris with your mistress, you were having an
operation, etc.). Simply the lack of evidence tells us nothing one way or
the other.

>>>(b) there is an absence of evidence that Fred's Bank on
>>>High Street was robbed by anyone yesterday.
>>In other words there is no positive evidence that proves OR disproves that
>>the bank was even robbed.
>No. Surveillance cameras do not report suspicious activities. No teller
>reports having been robbed. Records show that no money is missing,

Have you ever seen _Mission: Impossible_? You could have fed prerecorded
signals into the surveillance system; you could have replaced the stolen
money with counterfeit bills; you could have hacked into the computer system
and falsified the records; you could have staged a fire so the the missing
money would be presumed to have been burned; etc. Again, the lack of
positive evidence that the bank was robbed does not constitute positive
evidence that the bank was not robbed.

>>>Should this
>>>absence of evidence be considered evidence that I did not rob
>>>Fred's Bank on High Street yesterday?
>>No. At best it constitutes evidence that you both did and did not rob the
>>bank; at worst it constitutes no evidence at all. Either way, based on
>>this negative evidence alone no one can make a claim whether a robbery
>>occurred, much less whether you were the robber. In other words, the
>>evidence tells us that we can neither confirm nor deny that a robbery took
>>place and can neither confirm nor deny that the culprit was Brian Harper.
>>Practically speaking, however, the end result is the same as if the
>>absence of evidence were evidence of absence.
>>>Or, more in context,
>>>does the absence of evidence for a global flood indicate
>>>evidence that such a flood never occurred?
>>No, because a global flood is not refuted by a lack of positive evidence
>>that verifies it (i.e., negative evidence), but by the presence of
>>evidence that refutes it. Again, however, the practical results are the
>>same, hence the confusion.
>Really? I'm afraid I'm just not getting it. Suppose we take Last
>Thursdayism or Invisible Pink Unicornism?

Let me reiterate: there is positive evidence that refutes global flood
theory. Do you agree with that?

On the other hand, there is no positive evidence that either supports or
refutes the concepts of Last Thursdayism or Invisible Pink Unicornism. Do
you agree with that?

As such, we can say that global flood theory is refuted, but we cannot say
the same for Last Thursdayism or Invisible Pink Unicornism. However,
neither can we say that Last Thursdayism or Invisible Pink Unicornism has
been proven. We can in fact say nothing about them at all, so we can ignore
them as if they were not true, if we so wish to do.

Kevin L. O'Brien