Given the good news from Vienna this evening we can safely assume that
this Sunday it is going to be very difficult to find a place to have the
meeting. It is not every day that Spain gets to play at the final for
the Euro Cup. Having said that, as I write in my essay, we should always
be prudent when applying theories. Hence we'll see what is going to
happen on the day. Which brings me to the topic of what this Sunday's
meeting might be about: Do we have to trust theories?
Olga tells me that for this meeting she wants to learn more than she can
possibly contribute to the subject. Like you I always look forward to
her insights on the topics we discuss; luckily Sunday is still a long
way away. So may I recommend that she gets that Google search engine
working. In the meantime Olga wants to share the following short article
she found with you:
In methods we trust?
Volume 30, Issue 10, Oct 1997 Page(s):119 - 121
Digital Object Identifier 10.1109/2.625340
Summary:The degree to which you trust your environment-including
co-workers, software tools and systems-has a dramatic impact on the
performance of your entire software development team. If you can't trust
someone during a code review, what are the chances the code review will
be any good? I've become increasingly aware of another aspect of trust:
the trust of a method. A method is a disciplined approach to problem
solving that produces one or more well-defined outcomes. There appear to
be two philosophies when it comes to trust and methods. One camp soundly
rejects methods as a basis for trust: the only valid approach to problem
solving is to identify specific problems systematically in their
environment and solve these problems, one by one. The other camp takes
the opposite approach, performing every process step by preparing all of
the models (and other outcomes) defined by their method. The question is
not whether you build software according to a process, but whether you
trust the process you use. Do you trust your development process to
generate accurate schedules? Do you trust it to generate high-quality,
easily maintainable source code? Do you trust that your process will
generate highly usable systems?
Talking about methods we trust, Edwin has sent me the link to the BBC
website where he wrote about his experiences in May 1968. At least his
public domain experiences are very fascinating; highly recommend his essay:
This is the link to the bit I put on the BBC Website for 1968.
It was a good year, at least for me, there aren't that many in the
Tertulia who will have such strong memories of the year.
That year is part of the reason I'm here in Madrid now.
Take care and see you Sunday
IF YOU DON'T GET AN EMAIL BY FRIDAY PLEASE LET ME KNOW
Members' ads at the end of this message.
SUNDAY 6.00pm – 8.30pm at Molly Malone's Pub, probably downstairs----
-Yahoo group >> email@example.com <
-Old essays: www.geocities.com/philomadrid
- Blog: http://philomadrid.blogspot.com/
-Group photos: http://picasaweb.google.com/photosphilo
-My tel 606081813
-metro: Bilbao : buses: 21, 149, 147
Do we have to trust theories?
The epistemological status of theories is a central issue in the
philosophy of science. It is therefore not surprising that we find a
huge body of studies and research on the topic. The articles in
Wikipedia on Theory* and Scientific Method* are a good introduction to
our topic. So, I do not propose to give a historical account on the
philosophy of Theory.
An obvious answer to our question would be, it all depends on what
theories we are talking about. And if we decide not to trust theories,
we are still faced with what to do next. We can then be sceptical about
the replacement. But the obvious is not necessarily the best
philosophical investigation. One approach we can take is to analyse the
conceptual meaning of theory and then to see if this analysis is itself
sound and valid. The consequence of this analysis, I propose, would be
to consider the question, when does a belief become a theory?
The Wikipedia article on theory makes it clear that there is a big
difference between the common use of the term theory and the scientific
meaning of theory. I am particularly interested in the scientific
meaning of theory. We have more at stake in the context of science.
The question is then, what is science? Or what should we regard as
science? An other obvious answer would be to say what science
departments do in universities. A legitimate start, but a rather one
sided answer. Should we exclude politics, art history, ethics,
philosophy, business and industry? All these disciplines have their body
of theories, not to mention that every discipline has its own way of
research and investigating its subject matter.
Maybe we should broaden our definition of science, at the very least,
for our purposes. I would argue that a study of the nature and scope of
theory is to include any discipline that proposes to explain and
understand the world around us. For example, was Napoleon a
megalomaniac? Was he a short man? These are also facts about the world
not just topics in history. Maybe he was and maybe he wasn't
megalomaniac, but maybe no better or worse than those who came before
him. This is an empirical question which also depends on what we mean by
megalomaniac. As for being short, a Wikipedia article on Napoleon of
Popular Culture, suggests that this was just British propaganda at the
time and the fact is that he was of average height for the time.*
Going back to what we mean by theory, the Article on Theory suggest that
it is a formalised expression of our observation and which is also
predictive, testable and logical. One way we can understand logical here
is what the article on the Scientific Method points at: " any useful
hypothesis will enable prediction, by reasoning including deductive
reasoning." I do not propose to distinguish the difference between
theory and hypothesis, observation and experiment etc. What is relevant
for us is that logic can mean deductive logic (formal logic), inductive
reasoning based on statistical analysis, probabilistic reasoning and so on.
Ironically, the logic criteria for Theory is also the weakest criteria,
in my opinion. For example, deductive logic is silent about the veracity
of a proposition we put in an argument and deductive reasoning can be
abused by emphasising the importance of the methodology and not the
a) All martyrs go to heaven,
John is a martyr,
Therefore John has gone to heaven.
b) We should expect to see more miracles from a holy place with 67
recoded miracles over a period of 150 years.**
The problem here is that by talking about John and about holy places we
are really referring to facts about the world, but concepts such as
heaven and miracles are not facts about the world, but what we mean them
to be: your heaven is not as good as my heaven. That sort of problem.
However, both arguments meet the logic criteria. Of course, this does
not tell us much about heaven and miracles, but it does tell us a lot
about logic and its limitations.
I would say that predictability is the strongest feature of Theory
because of two basic reasons (there might be others of course). The
first is that theories can help us predict the future or describe how
past events were like in reality. The second is to answer one of the
most psychological preoccupation we have about life: planning for the
By being able to predict the future or how events really were, we
explicitly imply that the theory can stand on its own merits without
manipulation or fixing things. The ultimate proof of objectivity is to
verify predictions or falsify them. I do not wish to start a debate on
whether Popper was right or wrong here; and I am assuming that any one
can test a theory, but more about this later.
We are all familiar with the use of Newtonian physics to put satellites
in space. But this week there was a report*** in the New York Times
suggesting that a passage that appears in Homer's Odyssey might have
been referring to an eclipse that took place on April 16, 1178 B.C.
Scientists from the Rockefeller University were able to work backwards
in time the eclipses and celestial positions at the time of Homer. What
might have been regarded as metaphor might have been a description of a
fact in Homer. Not only can a good theory work backwards and forwards in
time, but can in this case give meaning to mythology.
Knowing that a theory does in deed work should give us a great deal of
peace of mind. Being able to predict the future also means that we have
the survival edge over most other creatures. Predictability means that
we can plan for the future, for example, our investments, pensions,
bearing children and so on.
Maybe it is because predictability is such a desirable feature that it
can easily lead us to abuse or misuse it. One of these misuses is to
assume that a theory predicts more than it actually does. For example,
Newtonian physics was once thought to be the answer to all questions
about nature from the nature of the mind to all things celestial.
However, it eventually became evident that Newtonian physics could not
explain everything especially the atomic world. But limitations do not
necessarily mean that a theory is wrong. Today, for example, Newtonian
physics and quantum mechanics can easily be employed side by side on the
If predictability is the benefit of theory, testability is the
combustion engine of theory. No matter how attractive a theory is, if it
cannot be tested it is not worth the paper it is written on. However, we
should be careful what we mean by testability. It is one thing not to
have the means or the tools to test a theory and another to have
untestable concepts. I do not need to give examples to illustrate the
first category, there are many examples easily available after a quick
search of theories that were tested years after they were proposed.
The example of martyrs going to heaven, is conceptually untestable. What
is a martyr? What is heaven? What would we have to do to test whether
martyrs do go to heaven?
However, the example about miracles taking place in holy places can
after a fashion be tested. But first we will have to disentangle the
semantic confusion from, say, medical science. Hence, it is not that in
these places there is regular divine intervention, but maybe in these
places the placebo effect is much stronger. Maybe because people go to
places because they have exhausted conventional medical care and
therefore there is a more urgent need to be cured. Maybe what the
faithful call a miracle are exceptional instances of the human body
curing itself. Something which is well know to those in science. Not to
mention the fact that the tunnelling effect in quantum mechanics is an
instance of the impossible taking place. We are therefore already
familiar with the idea of the impossible taking place in nature. In any
case it is irrelevant whether today's medical knowledge can explain such
cures. Maybe future knowledge might.
Another aspect of testability is that it does not matter who does the
testing. Anyone could have dropped the two different weights from the
tower of Pisa; it would not have mattered whether it was Galileo or not,
both would have fallen at the same speed. If, therefore, the person who
does the testing is also a key feature of the testing most likely we are
not taking about a scientific theory. One point of clarification, I do
not mean here the influence of the observer on the observed as described
by Heisenberg uncertainty principle. I am thinking more on the lines,
for example, that acts of parliament can only become laws if the monarch
gives the royal assent in constitutional monarchies. This is a
convention and not a fact about the world that can be described as a
theory. We cannot test a convention, and being able to distinguish
between the two is very relevant for out debate.
However, a serious problem with testability is that some theories cannot
be tested not because we do not have the mean or tools, but because
morality and ethics prevents us from carrying out the test. For example,
Levitt and Dubner in their book Freakonomics give details of Levitt's
study that showed that the fall in crime rates in New York during the
1990's was more the result of introducing abortion than any other crime
measures. However, we are prevented from testing this theory by having
two cities, one offering abortion services and the other not. Levitt et
al had to use historical data and it is unlikely that this could be
repeated in the form of an experiment.
Once we start analysing the aspects of theory we can find a number of
problems that might stop us from trusting theories. Theories have
limitations either at the testing process or at the prediction process;
I am assuming our observations are not flawed. However, the question "Do
we have to trust theories?" does need some philosophical
disentanglement. For example, what do we mean by trust. Does it mean not
relying on them or does it mean being prudent when using theories. In
fact the scientific method requires us a priori to be prudent when using
or testing theories. And as I have already pointed out, not trusting
theories does not put us in a better position despite their failings.
The problem for us, therefore, is not whether we should or should not
trust something that purports to be a theory, but rather:
a) Is a given theory is a scientific theory?
b) Is a given proposition a theory and or a belief?
If, for example, medical science has an impressive collection of
testable and provable theories does that mean that was is practiced as
medicine is based on a valid scientific theory?
Today we have no problems accepting the claims that unsafe sex increases
the probability of transmitting AIDS to others. And the claim that
smoking can cause lung cancer. Few people would dispute these claims and
those who do might be mixing up emotion (religious beliefs, business
interest) with science (multiple studies and tests).
Dan Ariely's study.+ on pain will make the point about distinguishing
theory from belief clearer. He concludes that the commonly held belief
within the medical community that removing bandages from raw skin or
flesh is best done quickly is basically false. The patient is not better
off if the bandages are removed the traditional way. But if they are
removed slowly but steadily this limits the amount of pain felt.
Ariely argues that this is a belief and not some valid medical
hypothesis since it has never been scientifically tested. The reason why
bandages are removed quickly is because the medical staff are genuinely
concerned about the pain and discomfort felt by patients. The intentions
are beyond reproach, but according to Ariely, it is not really science.
(Ariely himself was a serious burns victim and his experiences in
hospital moved him to study the issue.)
I submit that this is the real philosophical and scientific challenge
about theories. Sorting out theories from beliefs and prejudices fro
facts. And that prudence should replace trust and open mindedness should
I am therefore not about to stop believing in Newtonian physics and will
certainly keep on believing that the sun will rise tomorrow. What I am
not sure about is what to do next time I burn my hand on a hot stove
when I am preparing diner. My theory is to play safe and let others do
the cooking. Not only have my observations been that I never burn myself
when others do the cooking, but the bonus is that their cooking is
always much better than my efforts. A priori, I can therefore predict
with certainty that if I do not man handle hot pots and pans I reduce
the risk of burning myself in the kitchen. And in the spirit of the
scientific method I will let others test the theory that cooking can
cause burns to the skin.
Napoleon in popular culture
***June 24, 2008
Homecoming of Odysseus May Have Been in Eclipse
By JOHN NOBLE WILFORD,
Copyright 2008 The New York Times Company
+Predictably Irrational: The Hidden Forces That Shape Our Decisions
By Dan Ariely
Mayte; Almería (Villa de Níjar);
Paloma; Marbella (near Elviria);
from Lawrence, Pub Philosophy Group, Sunday meeting: Do we have to trust