It's not my name for it."the
Principle of Verification" that is. You can blame the Vienna
Circle and, particularly, A
J Ayer for that mouthful. But it is a pretty simple concept to wrap
our heads around. In short what they're saying is that if you've got
an idea about the world the test of its "significance" is that
you should, in principle, be able either to prove or disprove, i.e.
to verify it. A couple of examples might help. First, an obviously "unreasonable"
proposition:- "There is a fairy standing on your left shoulder. She
is only visible, however, when no one is looking at her. She is made
of mist; so she weighs nothing and you can't feel her in any other way.
Any other attempt to record her presence makes her disappear instantly."
Now then, how are you going to deal with this proposition? You suspect
that it is untrue of course. However, you can't prove it to be untrue
because we have deprived you of any means of doing so. You might, for
example, have had the idea of getting yourself recorded on camera to see
if she turns up on screen. But our last sentence rules that out as any
such attempt will cause her to disappear in any case.
The only sensible way to treat such a proposition is to ignore it.
It is effectively "meaningless" as you can not even imagine how to ascertain
whether it is true or false.
What about this one though. "The Moon is made of Mature Cheddar". Believe
it or not, to a philosopher, this is a "meaningful" proposition! The reason
is simple. We can test it. We merely catch the next passing Apollo, grab
a lump of moon and taste it. This establishes pretty quickly, as we all
now know, that the proposition is untrue. Gorgonzola perhaps. Cheddar
- you must be kidding.
The point is that whether or not a proposition is meaningful depends
not on whether it is true but only on whether you can establish
its truth or falsehood. In other words you
can have perfectly meaningful propositions which are entirely untrue.
We chose the Moon as an example for one particular reason. Which is that,
in practice of course, very few of us would be able to test the proposition
regarding its makeup. It is vitally important to understand that this
does not weaken its "meaningfulness"
. What matters is that it is testable in principle. i.e. We can
design experiments or procedures which would test the proposition if we
had the resources and technology. Hence, for instance, all the following
propositions are meaningful - whether or not they are true - simply because
we can do something to establish their "Truth
Value". With that in mind, here are some more...
* Grass is blue.
* Elephants are afraid of mice.
* There is a Tenth Planet in our Solar System.
* Men are more intelligent than Women.
* The only way to disarm Iraq is through military intervention
Some carefully chosen examples there. All "meaningful" because we can
test them. The choice illustrates a couple of useful points which you've
probably already noticed. First, its not always easy to test a proposition.
OK, the grass one is easy. Just
look at it. But how exactly are we going to test whether Elephants
fear mice? We don't know what is going on in an Elephant's head, so even
observation of its behaviour won't necessarily inform us one way or the
other. We might, perhaps, design an experiment in which mice are let loose
in the vicinity of a random selection of Elephants and watch what they
do. If they turn and run, we might deduce that this demonstrates fear.
It is possible, however, that the gentle giants simply don't want to hurt
the wee rodents so they gallantly vacate the vicinity. Conversely, apparent
indifference need not indicate the absence of fear. It may be that they
are so petrified that they are frozen into immobility. You get the gist.
No doubt, with a certain amount of planning and forethought, one could
actually design a credible experiment which gave a reasonably reliable
pointer to the truth. It wouldn't be easy though and you may have to concede
a lingering doubt about the interpretation of any results.
By contrast, investigating the proposition regarding a tenth planet is
much simpler in principle though somewhat more difficult and expensive
in practice. You just knock up a spaceship and go look for the damn thing.
You might even solve the problem with a telescope search. If you find
it, that's the end of the argument. The problem with this one is that
if you don't find it, you haven't necessarily proved that it doesn't
exist. Its truth value is unresolved. This has led some philosophers to
argue that such propositions are - like the fairy - also meaningless;
because you can't design an experiment to refute the hypothesis: Karl
Popper even described
Evolutionary Theory as Metaphysics because it was Tautological
and, thus, not Falsifiable. In addition to the erudite arguments you will
find under that Evolution link, we would add two further answers to this
To begin with, the argument really reveals a failure of the imagination.
OK, so the kind of exercise we could mount today might be capable of "proving
the positive" - by finding the planet - but the task of proving that there
is no such planet is clearly well beyond our present or even prospective
capabilities. We would need to be able to establish beyond doubt the precise
contents of a sphere in space whose diameter was perhaps 10 times the
present Sun-Pluto orbital diameter in order to be certain that there was
no such planet. This would be roughly the equivalent of checking every
grain of sand on this planet in order to be sure that one of them was
not a small diamond. It sounds impossible and it may well always
be beyond our reach in practice. But it isn't impossible in principle.
There is no known physical law which would prevent us from successfully
conducting the search. It is merely too difficult for us to handle for
the time being at least. It is also apparently pointless. The reward for
such efforts would not compensate us for the huge cost of conducting the
search. So we don't and won't bother. But that is still not an objection
in principle, it is merely a practical objection.
One hundred years ago, they could have raised the same objection to most
propositions about the moon. They had no way of testing those either.
But the barrier was technical, not logical. Who knows, we may eventually
develop space technology which will be more than adequate to the task.
Indeed, there might be other species out there whose technology already
can prove the non existence of the tenth planet. In which case, of course,
the proposition is clearly meaningful to them. To argue that it can't
be meaningful to us simply because we haven''t yet got the technology
to prove it doesn't exist is plainly nonsense.
The second argument is simpler. We can easily imagine proving the positive
and, if we did, then not only must the proposition be meaningful, it would
also obviously be true. Yet to argue that a proposition could be meaningful
only if true is to miss the whole point which we discussed earlier.
The essential point is that for the proposition to have meaning we must
be able - in principle even if not in practice - to prove either
that it is true or that it is false. The tenth planet problem
is thus not so difficult for the philosopher after all. It might be difficult
or even impossible to prove it false. But finding the planet would be
a relatively simple way of proving it true. This is enough to confirm
that it is philosophically meaningful.
As we've mentioned, there is a famous example
of this argument in which Karl Popper, who prefers a principle
of falsifiability (local)
argued that modern theories of Genesis and of Evolution are not
Philosophically meaningful and are mere 'Metaphysical Research Programmes'
because we can't prove that they are false. (To be fair, Popper later
his objections, which just goes to show that even world famous philosophers
can make - and admit - their own mistakes) Again, his argument was based
on technical rather than logical objections. i.e. He couldn't conceive
of a way to test the theory. Well, in the first place, what would we
need to do to establish either its truth or falsehood? Presumably, we
would have to find a spare sterile planet and set up an experiment in
which conditions were similar to those we expect in the early stages
of a life-bearing planet. We then sit around for a few billion years
and see what happens. While we're at it, we'd better set up several
hundred such experiments together with controls in which the conditions
are different in order to be statistically confident that we are seeing
real effects and not "artifact" due to our contamination or whatever.
If life in some shape or form sprang from these experiments and begins
to evolve, then clearly the theory would be validated(ish).
Moreover, the theory could 'easily' be shown to be false by proving true
some rival contradictory theory. (eg what the creationists
allege - that all life and the fossil record spring into being both spontaneously
and simultaneously.) The fact that the experiment is at least conceivable
- there are no known physical laws which would prevent it being carried
out - is enough to make the proposition /theory meaningful. OK, we can't
-yet -even reach a suitable sterile planet, let alone "terraform" a few
hundred of them. Neither have we the time to wait a few billion years
for the results. But these are the usual trifling practical objections,
not fundamental logical blocks. Elsewhere or elsewhen, it might be a trivial
Of course, one might reasonably object that even experimental success
of this type would show only what took place in those instances. But this
is no more than repetition of the points we've already agreed in discussing
of perception. (and is also an example of the so-called "problem
of induction") We have accepted that limitation and decided that what
we have to settle for in place of knowledge is rational understanding.
Theories such as evolution are clearly able to advance such understanding
and present rational arguments for the phenomenon of diverse life.
You see the point. Popper was merely guilty of a failure of the imagination.
There is no logical reason why we could not set up such an experiment
- even if there are one or two practical difficulties. The "ish" after
'validated' is there because, in practice, such an experiment could lead
to results capable of different interpretations in similar fashion to
the Elephant and Mice problem. For example, whilst it could demonstrate
fairly clearly that life will emerge from inorganic chemicals given time
and conditions, it may not support so conclusively the main tenet of Darwinism
- that species evolve through a process of natural selection. The evidence
might lend weight to other theories at the expense of natural selection.
This is not a problem for the philosopher. The mere fact that we can't
even imagine what shape such alternative theories might take is irrelevant.
It just shows the limits of our imagination. As philosophers we
don't care, in a sense, what the truth is. What matters
to us is whether you can - in principle - determine what it is.
We can possibly also help in deciding how to test it.
Lets just return to the missing planet problem for a moment to demonstrate
the major weakness in too literal an interpretation of "falsifiability".
Essentially the view of the stringent falsifiers is that as no hypothesis
can be "proved" by the examples of instances which support it but can
by instances which don't, then the emphasis should be on the search
for those disproving examples rather than the supporting cases. Validation
then rests upon your continued failure to find such examples. The harder
you look and the more failures you have, the more valid the hypothesis
is seen to become. This may sound unnecessarily tortuous. To give a simple
example where falsification makes sense, lets say we hypothesise that
"all swans are white". We may have reached that 'provisional conclusion'
on the basis of seeing a few swans. The falsifier approach is that simply
finding more white swans is not enough to validate our hypothesis. What
we must do is actively try to find a black one (or any other colour).
And it is our continued failure to find one that continually increases
our trust in the hypothesis. Like verifiers, falsifiers also understand
that continued falsification never finally validates our hypothesis.The
first non white swan and 'Poof'! the hypothesis will sink without trace.
Falsifiers, in this sense, can be seen as no more than "fundamentalist
verifiers". A verifier basically says that after a reasonable period
of verification, although we'll never be certain, it is rational to accept
the apparent validity of the hypothesis, on the understanding that should
contra indications ever present themselves, we may need to revise or ditch
the hypothesis. The falsifier is merely obsessive about maintaining the
level of doubt.
But now lets consider the example of a real 'missing planet' hypothesis;
viz the famous one which led to the discovery of the planet Pluto. This,
we learned in elementary school, arose from observed perturbations
in the orbit of Uranus and Neptune. The hypothesis was that these were
being caused by a so far undiscovered 9th planet whose orbit must, at
some point, intersect or come very close to that of Neptune. On the
basis of this hypothesis the prospective planet was searched for in
approximately the right place at approximately the right time and as
we all now know, was found to be exactly where the hypothesis said it
should have been. And the ironic beauty of this story? Its not true!
Or rather it is true but it shouldn't have been. It turns out that Pluto
is far too small to account for the observed discrepancies in the apparent
orbits of Neptune or Uranus. Essentially, estimates of planetary mass
- and thus their gravititational effect - were significantly mistaken
and when we correct the error, there is in fact no
unexplained perturbation. Yet we still found
Pluto - right where our mistake caused us to believe it should be!
(Now there's the seeds of a cosmic conspiracy theory if ever we saw
Now, what happened to the scientific hypothesis in this situation. From
the point at which EW
Brown revealed that the calculations were wrong, astronomers realised
the hypothesis had failed. Nevertheless, search text books up through
the 1950 and 60s and the hypothesis is still offered as the explanation
for the discovery of Pluto, despite Brown's published - and accepted -
debunking. In fact, right up until our satellites were able to get close
enough to Neptune and Uranus to measure their masses accurately, it was
still widely considered an impressively validated hypothesis. Obviously
we should have listened, on this occasion, to the warnings of the falsifiers,
under whose rules we were, of course, never entitled to conclude that
it was validated. They're now able to say "we told you so" (metaphorically,
that is; no-one objected at the time) The theory was falsified by the
revelation that the initial data - which fuelled the assumptions about
the relevant masses - was wrong. The fact that we discovered a planet
where we happened to be looking, for entirely the wrong reasons, was just
an amazing coincidence. (Perhaps)
But what if the data had been confirmed by multiple satellite
and other experiments? and Pluto proved sufficiently massive to account
for the perturbations? Then clearly we would still believe the hypothesis
to date. The question for the falsifiers is what would they offer by
way of a means of falsifying such a well supported hypothesis? To which
the fair answer would be along the lines of "we don't necessarily
know - but keep looking, just in case" and verifiers wouldn't argue
too much with that. The relevant question, for ordinary people, is "when
is it rational to trust the hypothesis?" Probably our best guide
to that is "Occam's
Razor". The simplest answer, consistent with all the evidence, is
to be preferred in favour of more complex explanations which do not
offer greater explanatory or predictive power.
The simplest explanation in the light of the evidence was clearly that
Pluto was the cause of an apparent anomaly. It appeared to fit the observed
data so well that it was considered unnecessary and foolish to search
for further explanation. It was, during that period of ignorance - based
on wrong data - rational to believe that Pluto was the cause of the effect
we believed we perceived. The fact that it turned out that this rational
belief was entirely misplaced teaches us a valuable lesson about ALL rational
beliefs. They are all rational based on what we know now. It is conceivable
in ALL cases that evidence will arise which can undermine each and every
scientific theory we have and destroy the basis for every rational belief
we hold. Ultimately, although one's gut feeling may be that "The
Matrix" is pure fantasy, and there are serious inconsistencies
in the story, there is no logical reason or physical evidence that prevents
its basic premise (that everything we perceive is essentially a Virtual
Reality projection created by something more advanced than we are)
being close to reality.
The key thing is that neither verifiers nor falsifiers have any difficulty
incorporating the corrected data into their world view. They adjust the
world view accordingly. In other words, there really isn't a conflict
between the principles of verifiability and falsifiability. All rational
and meaningful hypotheses must surely meet a mix of both criteria. That
mix will be determined, pragmatically, by the number of observations it
is possible to make of the phenomenon we are trying to explain. Finding
a new planet can, by definition, be satisfied by a single observation.
Having found it, its a fact. You don't invalidate a planet by 'not' finding
it in other places!! Or even by discovering, as with Pluto, that you shouldn't
have found it in the first place! And even if the damn thing disappears
one day, it won't change the fact that it was once there - at least to
the extent that we are currently capable of determining such data.
Consider, though, the hypothesis that a diet rich in animal fat is unhealthy
for most human beings. Unlike a planet, where a single chance observation
leads quickly to confirmation of its existence, measuring something like
the effects of a diet high in animal fat requires thousands, possibly
millions, of observations before any hypotheses can be truly firmed up,
so here it makes much more sense to look for negative results like evidence
that some individuals thrive on the stuff. Or that the afflictions attributed
to high fat diets can in fact be laid at the door of other dietary factors.
The Theory of Evolution is clearly a mix. As a proposed chain of events,
it is, on this planet at least, apparently a 'one off' (to date at least)
and thus falls on the 'verifiability' side of the coin. Yet any 'sub-hypothesis'
involving the precise mechanisms and specific evolutionary pathways of
given species requires multiple observations and is thus more appropriately
subject to the test of falsifiability. It seems to come down to class
size! If the 'class' is small enough and you are thus able to measure
the entire 'population' (i.e. one way or another you can record every
instance of an event), then verifiability makes perfect sense. If you
can only expect to see a 'sample' of the events, then we need to use falsifiability.
The discovery of Pluto also illustrates what we might call the maximum
possible success of the empirical method.
Though we now know it is not having any significant influence on the
orbits of either Uranus or Neptune, we still discovered it and the reality
of its existence has thus outlasted the validity of the hypothesis which
led to its discovery. It was not part of the truth for the phenomenon
we were trying to explain (the apparent orbital discrepancies) but it
seems to be established in its own right as part of the wider truth.
What was the next one? Oh yes, the superior intelligence of Men over
Women. Mainly, of course, we include this one in order to invoke routine
prejudices. Hardly anyone deals with this sort of proposition dispassionately.
Male chauvinists argue
that the evidence is overwhelming and only political correctness prevents
it being accepted in the public domain. Militant Feminists, on the other
hand would probably organise hate mail against any scientist who dared
to publish serious research which supported the proposition. (Remember
about the research which suggests that a woman's brain shrinks (local)
during pregnancy? ) The philosopher has to take the objective view in
order to ascertain the truth. Whether and why s/he chooses to try to
impart the consequences of such views to society at large is a matter
of value judgement which we will get on to under the Third
The prime question for the philosopher is to decide is whether the proposition
is meaningful and only if it is would s/he go on to decide what it meant.
i.e. can it be tested and, if so how? And the answer in this particular
case is pretty clearly that it can be tested; by first establishing our
criteria for intelligence and secondly by examining a sufficiently large
sample (all, if necessary) of both sexes and seeing how well they match
the agreed criteria. That would settle the beyond dispute if we ever really
need to know the answer. The questions as to what use you make of such
information are again Third Question discussions and we'll probably come
back to them later; though we will take this opportunity to plug a most
intelligent discussion on this and related issues which is the book by
Janet Radcliffe-Richards called "The
Finally we come to Iraq.
The proposition is that: "The
only way to disarm Iraq is through military intervention".
The first point to make is that it is a valid proposition. It is meaningful
because we can imagine credible means by which we could either verify
or falsify it. For example some argue that the
only way to disarm Iraq is to remove Saddam from power (local)
and although military intervention is a practical means of achieving
that end, this version of the proposition doesn't "require"
military action. It merely "requires" Saddam's removal. Hence,
if the proponents of this view have an alternative means of disposing
of Saddam, and this were to result in the disarmament of Iraq, then
the initial proposition - that military intervention is required - would
be proved wrong.
Others argue that the
only way to disarm Iraq would be to reinstate some of the inspectors'
earlier freedoms. (broken - looking for new link) Here there is
not even any mention of regime change; while here:
only way to disarm Iraq is to Threaten the use of American Military
Force (broken - looking for new link) the implication is that the
threat is the important thing.
Essentially we have at least four apparently valid propositions about
how to deal with a perceived problem. How can philosophers help with a
problem like this? By clarifying the issues and options. And why should
they help? Karl Marx answered that best:
philosophers have only interpreted the world in various ways; the point
is to change it. (local)
The first problem with any of the propositions is the phrase "the
only way". As it is easy to conceive of alternative methods, all
four propositions are essentially falsified at the first hurdle (see how
easy it is to debunk a political statement?). However, we know that the
proponents are not arguing with philosophical rigour, so we'll cut them
some slack. What they obviously mean is that their particular proposal
is "the best" rather than "the only". So lets tackle
the problem from that starting point.
Like Pluto's discovery, one possibility is that the data we believe is
pointing to a particular conclusion is incomplete or invalid. Iraq may
simply not be the threat alleged. That requires us to gather the best
and most complete evidence we can, which is what makes UN inspections
imperative. Thats the first philosophical guidance. Until we know what
we're dealing with, we shouldn't make any rash decisions. See how close
it is to common sense? (Always a good test in philosophy).
What we do next depends on what the evidence reveals. The possibilities?
1 That Iraq is clean and innocent. It accounts for the destruction of
all offending weapons and materials. Appropriate action? Apologise, lift
sanctions, bring troops home.
2 That Iraq is guilty. It does possess the offending weapons but it co-operates
in revealing and destroying them in an obviously open and honest fashion
much like the South
Africans did, in the last days of the Apartheid regime. Action? Verify
destruction, lift sanctions, bring troops home.
3 That Iraq is guilty, but limits co-operation to doing as little as
it believes possible to satisfy the international community that it is
in fact behaving in line with the previous option while the inspectors
confirm that they are not getting South African levels of access or co-operation.
Action - persuade them that unless they do rapidly switch to the South
African model, steps will be taken to ensure that the weapons can be found
and safely disposed of, with or without Iraqi co-operation. These steps
include various means of replacing the regime with one more likely to
be compliant with international demands.
4 Iraq is shown to possess or not to have accounted for known weapons
which are in breach of UN resolutions and refuses to acknowledge the conclusions
reached by the inspection process or to cooperate in removing the weapons.
Action - same as (3) but with less sympathy, shorter timetables.
The most likely option at the time of writing (late Feb 2003) is the
third one. The implication is that regime change will be a necessity.
The next obvious question is how that will be implemented. The options
Bribery / Blackmail:
Covert "surgical" targeted assassinations
Full scale invasion and occupation of the country.
These options are discussed in more polemical style here.
The only point of including the topic in this chapter is to emphasise
what we have said before, in regard to the relationship between "Truth"
and the significance or "meaningfulness" of a proposition. There
isn't one. The points of view expressed in regard to the Iraq problem
are all meaningful propositions. Clearly they can't all be true. Indeed,
their use of "the only way" is likely to have rendered them
all logically false.
By now it should be clear that the keywords "Logical" and "Meaningful"
have a very precise meaning in philosophical discussion which is not quite
the same as their colloquial use. Let us ram the point home. In ordinary
usage, "meaningful" is usually taken as the equivalent of "comprehensible".
i.e. Anything you can "understand" is considered to be meaningful. In
philosophy we must be more precise and distinguish between mere comprehensibility
and genuine meaning. "Meaningful" in this context is much closer
to "Significant" than "Comprehensible". The philosopher
insists, reasonably, that any meaningful proposition must be either true
or false. If it is impossible to conceive a way of testing whether a proposition
is true or false, then it simply has no meaning - even if, like the fairy
story, it is entirely comprehensible.
Similarly, "Logical" nowadays has the same connotation as "Rational".
Here, we must distinguish between a condition of the mind and a condition
in language. It is, for example, both rational and logical to assert that
2+2 = 4. This is merely a rule in our language which helps to define the
term "4". It follows logically from propositions like "2 = 1+1"; "3 =
2+1" and "4 = 3+1". It is perfectly rational to recognise that.
However, it is also perfectly rational to avoid walking across a busy
main road without checking on whether there is any traffic approaching.
But it has nothing to do with "logic". i.e. there is no "rule" which states
that you must be knocked down every time you cross such a road. Chances
are that nine times out of ten, you wouldn't be knocked down. But that
'one in ten' chance is too high a risk. Clearly, however, it is not a
certainty which follows from the definition of "busy road" in the
same way that "4" follows from its prior definitions. Our reaction is
based on experience in this case rather than predefined or immutable laws.
Conversely, it is "logically" possible to assert a false proposition
- such as "3+2 = 4" but it would not be "rational" to agree with it! This
is because logical propositions merely have to be the correct "form";
we must be able to analyse them. They don't have to be true. That is for
the analysis to decide. Rational propositions don't have to be true
either - but they must at least accord with the evidence available to
the person making the proposition. For instance, to those born in
earlier centuries, the proposition that the Earth was flat was entirely
rational because they had no satisfactory evidence against it. We believe
now, of course, that the same proposition, if made today, would be entirely
irrational as there is somewhat compelling evidence to the contrary. Logic
doesn't shift ground like that. Both the propositions that the Earth is
flat and that the Earth is Spherical are equally logical. The fact that
they can't both be true does not weaken their logic.
Another way of looking at the difference is to recognise "Rationality"
as the process of deciding (amongst other things) which "Logical" propositions
are "True" and then acting accordingly.
Well if all that's clear, its time to take a look at How
We Got Here...
(Last Updated 23 Feb 2003)