ArticlePDF Available

Philosophy and methodology of military intelligence: Correspondence with Paul Feyerabend



Without Abstract
Philosophy and Methodology of Military Intelligence:
Correspondence with Paul Feyerabend
Philosophy and Methodology of Military intelligence - Correspondence with Paul
Feyerabend, Philosophia, Vol. 28, Nos. 1-4, pp. 71-102, June 2001
In 1989 I wrote a paper on the logic of estimate process in military
intelligence1. Since its central ideas were drawn from philosophy of
science, I sent a draft to Professor Paul Feyerabend. I knew that my paper
will intrigue him: after all, he is the one who wrote a book Against
Method2 in science, and here I come and dare to claim that military
intelligence would gain a lot by adopting the scientific method!
Surprisingly enough, I have got a (relatively) positive response from
him, and this had started a chain of letters, which lasted for some years.
The main correspondence is given below.
Before going to the correspondence itself, I think it would be
beneficial for the reader to summarize, in a nutshell, the original paper I
sent to Feyerabend.
The Logic of Estimate Process - A summary
My starting point was Israel’s intelligence failure in predicting the
outbreak of war on 6 October 1973. Many believe that it was a result of a
certain fixed notion (‘The Concept’), universally and rigidly held within the
military establishment: Egypt would not go to war without a long-range air
strike capability against Israeli airfields; Syria would not go to war alone;
Since Egypt had no such capability, the probability of war was believed to
be very low.
Some consider the general acceptance of this Concept as the root of evil.
The fact that Israel’s intelligence heads had an a priori concept about the
necessary preconditions for outbreak of hostilities is considered to underlie
the wrong estimate. A good intelligence estimator, they say, must free
himself of all commitment to any single conceptual framework. Others
claim that an intelligence estimate is not possible without some kind of
conceptual framework.
1 I. Ben-Israel, "Philosophy and Methodology of Intelligence: The Logic of Estimate
Process", Intelligence and National Security, Vol. 4 No. 4, October 1989, pp. 660-
2 P. Feyerabend, Against Method, New Left Books, 1977.
Should one aim to eliminate conceptual framework? Can this be
done? Is it possible to make an intelligence estimate without a conceptual
framework? Does such a framework have any ‘positive’ role? If so, how
should a conceptual framework in intelligence be built? What risks lurk
within it? How and when should it be dropped? All these questions connect
with one central question: is it possible to indicate methods for intelligence
estimates which are ‘better’ and more ‘successful’ than others? Or, to
formulate it more bluntly: how should one carry out an intelligence
Similar questions may be asked, in fact, in almost every field where
information is gathered under conditions of uncertainty, processed, and
used for forecasting. Intelligence is nothing more than an institution for
studying and clarifying reality, and hence there is a clear analogy between
intelligence and science (which is also such an institution with the same
The intelligence field has its own particular aspects: it usually
involves risking human life, so the cost of error is extremely high; security
problems dominate and most conventional techniques for filtering errors
are often blocked by security restrictions. Not only is intelligence material
itself considered classified, but also its method of work. The classification
of method is a serious obstacle to the development of intelligence, since it
prevents open, systematic discussion of the methodological question of
‘how the intelligence estimate should be done’ (as well as many other
related questions, like those mentioned earlier concerning conceptual
For these reasons, I chose to conduct my analysis in what might
initially be viewed as a devious, roundabout way: by studying conclusions
and results accumulated during centuries of research in an entirely different
field – the field of philosophy of science.
I began the paper I sent to professor Feyerabend by challenging the
traditional method used for intelligence estimate, contending that a method,
which inductively derives its conclusions from known data, is wrong.
Next, I claimed that the ‘business’ of intelligence estimate is to derive
predictions from information. This is also the case in science. There is thus
an analogy between science and intelligence at the level of the logic of
prediction. I then proposed an alternative critical method, based on
Popperian doctrine in philosophy of science.
This method is far better than the traditional approach, but not quite
good enough. I therefore suggested a number of modifications and
amendments. The resulting amended critical method turns the process of
estimate upside down: conjectures (hypotheses) must first be raised and
only then can ‘facts’ be approached – and even then, not to verify the
‘chosen’ estimate, but to refute the competing ones.
There remains, however, a substantial difference between ‘science’
(dealing with ‘dead’ and passive matter) and ‘intelligence’ (whose research
object is active human beings and societies with free will). This difference
prevents the application of specific scientific categories and methods to
intelligence (or to any other branch of social science). In researching
people and societies, it is never possible to transcend the realm of
conjecture and hypothesis.
Nevertheless, I showed that my proposed amended critical ‘method’
(the logic of research), which considers all science as a set of conjectures,
is applicable, in principle, to the field of intelligence estimate.
Furthermore, I demonstrated that the practical difficulties in applying
my method to intelligence could be overcome. In confronting these
difficulties, some practical methodological rules were drawn.
I also showed that conventional alternative methods, such as
historicism, are not valid for intelligence estimate, neither in pro – or
antiscientific form.
I concluded the paper I sent to professor Feyerabend by analyzing a
major historical example: the intelligence failure of the October 1973 war,
in order to clarify the main differences between the various methods.
26 September 1988
To: Professor Paul Feyerabend,
University of California, Berkeley
Dear Sir,
I am sending enclosed a copy of my paper, “Philosophy and
Methodology of Intelligence: The Logic of Estimate Process”, in the hope
that you will find it interesting enough for reading and sending me your
The basic idea of the paper is a very simple one. Since I consider
philosophy of science to be valid for any attempt to construct a systematic
epistemic knowledge, I do believe that it can be applied to any field of the
so called social sciences, and in particular to the field of military
At first sight, it might seem as if I claim something that is in the
opposite spirit of Against Method. Well, I do believe that it is not so. If I
read you correctly, you are not against all methods whatsoever, but against
a single ruling one.
Philosophy and history of science teach us that there is not an a priori
‘right’ method for tackling a specific problem. It doesn’t teach us that
methods are totally dispensable.
I tried to explicate these ideas, as clearly as I can, in the enclosed
paper. In a way, it is a ‘case study’ in one particular (underdeveloped) field
of ‘science’.
The paper was accepted for publication in Intelligence and National
Security, which is not a journal for professional philosophers, and therefore
I had sometimes, to go into otherwise oversimplifications.
I shall be grateful for any critical comment you care to make.
Looking forward to hear from you,
Yours sincerely,
Dr. Isaac Ben-Israel
The Institute for History and Philosophy
Of Science and Ideas,
Tel-Aviv University
30 November 1988
Dear Isaac Ben-Israel,
Thank you very much for your fascinating study of the intelligence
process. The suggestions you make are eminently reasonable, you have
recognized that there are bound to be conflicting requirements and have
recognized the need for a new kind of mathematics (your p. 74 – let me add
here that foundational studies in quantum mechanics have moved away
from differential equations into lattice theory and algebra, the direction
suggested by you). But I wonder if the detour through the philosophy of
science was really necessary.
To start with, the critical method was not “born in the philosophy of
science” (p. 31), it is old hat; it was used by the discoverers of new
continents, by businessmen like Marco Polo, by Generals like Clausewitz
(whom you quote) and it was matter of course for the native tribes in
Kenya (in the thirties) who, being faced with invaders of the most varied
kind became more critical than the local missionaries who met only other
Secondly, this ‘naively’ (i.e. unacademic) critical approach will most
likely be more effective for it can also work in partly closed surroundings.
Businessmen were and still are rivals, there can be a flow of information in
some direction, not in others and so criticism here is adapted to conditions
necessary for intelligence work but absent (to some extent!) from the
sciences [I think it is an illusion to expect an opening of intelligence work.
This not only will not occur, it must not occur as long as the present
political situation prevails]3.
Thirdly, there is now the question of implementation, the most
important question. In second-world-war England an improvement of
intelligence work was achieved not by reforming the intelligence
establishment via theory, but by introducing a second intelligence network,
consisting of laypeople (in the field of intelligence) entirely: actors (like
Noel Coward and Leslie Howard), scientists (like Turing) and others who
had never done any intelligence work and were therefore unconfined by
intelligence-prejudices (except the prejudice of secrecy). I may be wrong,
but such a replacement seems to me more effective than a new training
for the old cadres – and so, again, philosophy of science can be dropped
from the scene. Besides, it does not even help within the sciences, which
proceed in their own irrational way from discovery to discovery.
3 Feyerabend added this note in the margins of his letter.
So, in sum, I find your arguments excellent, your conclusions plausible
– within the framework you have set yourself but I find this framework
itself more a hindrance than a help. This, of course, may be my own
sizeable personal prejudice. At any rate, thanks for sending me an
interesting piece of material.
Paul Feyerabend
1 January 1989
To: Professor Paul Feyerabend,
University of California, Berkeley
Dear Sir,
Thank you very much for your kind letter concerning my paper on the
logic of intelligence process. I am really grateful for the time you spent on
reading it and especially for the critical comments you cared to make.
Your main criticism concerns the ‘link’ between intelligence process
and philosophy of science: you take it to be unnecessary and “more a
hindrance than a help”. If I read you correctly, you mean that (A) my
analysis of the intelligence process and my suggestions are independent of
philosophy of science, and (B) philosophy of science does not even help
within the sciences.
Well, I accept both (A) and (B) above: science (as well as
intelligence!) is a practice, and it does not depend at all on philosophy. If
philosophy could improve science, it would probably be compulsory for
students of science (and it is not)! Philosophy does have some effect on
science, but it is a very indirect one through the influence on the cultural
climate (with science as one of its manifestations) and not on the internal
progress of some particular field of science. It is a well-known fact that the
so-called ‘revolutions’ of the 17th and the 20th centuries were not confined
to ‘science’ only. They included the arts, literature and poetry, architecture,
military practice, social constructs, etc. Science is only one field of humane
civilization, and it cannot be conducted in ‘free space’, isolated from its
Query: are there societies and cultural climates that are ‘better’ for
scientific progress? I believe the answer is in the affirmative, though I
agree that these ‘better’ climates are not sufficient for scientific progress/
Moreover, they are not even necessary. For example, I do believe that
‘openness’ (glasnost!) is better for scientific progress, but I know that it is
neither sufficient nor necessary for it.
In a way, the relation of philosophy to the practice of science (as well
as to intelligence) is analogous to the relation of an elevator to the practice
of reaching an apartment at the top of a very high building: the elevator is
neither sufficient (it wouldn’t help without electricity and a key), nor
necessary (one can still use the staircase); Nevertheless, it helps a lot!
The remarks above may seem a little confused, indecisive and
ambiguous. But this is also the face of ‘reality’. Therefore, our efforts to
understand Nature (the scientific effort) cannot be described entirely in
‘rational’ and ‘logical’ terms. This is not to say that we are not allowed to
inject some order (‘logic’) into that mess. I think that this is what we are
doing in the so-called ‘philosophy of science’, and this is what I tried to do
in my paper.
I agree with you completely about the origin of ‘critical method’: it
was not born in the philosophy of science (and I shall correct my paper
accordingly). Nevertheless, I had my own reasons for introducing it
through the framework of philosophy of science. Let me explain it in some
To begin with, intelligence theory is in a very poor state. It almost
doesn’t exist. Now, not every practice needs a theory. But intelligence is
(as the very term itself hints) a highly theory-laden intellectual activity.
Therefore, I believe there is place for a revolution here. Unfortunately there
are not many supporters of this view.
Some of the opponents (I shall call them group-A) do not understand
the role of theory in their activity. They regard it as simply observing the
enemy and deducing some conclusions from observational facts. Others
(group-B), realizing the complexity of reality, do not believe in the human
ability to arrange it in a systematic way.
I use philosophy of science as a weapon against these two groups. To
the first group I say something like this: “don’t be so naive. Your
‘observations’ are theory-laden. You must admit it in order to fight your
prejudices (‘false idols’). Unless you do that, you will remain in the same
backward state as was science before Bacon and Newton”. In other words,
I use science and philosophy of science against group-A, as a tool for
propaganda. I want them to convert to my view, and I use the highly
respected phenomenon of science as an example. An example, of course, is
not a proof. But I want to convince, not to prove.
This high respect for science stands at the root of the view of group-B.
Science, they say, is a rational activity, which is successful because it deals
with dead matter; Intelligence, on the other hand, is totally different: it
deals with human irrational nature, and it cannot be done in a rational way.
Well, against this group I use philosophy of science in order to
demonstrate the falsity of their view about science: it is not so ‘clean’ and
rational, as they believe. And if science doesn’t proceed in a rational way,
then the basis for their contention drops. Again, philosophy of science does
not prove that a similar progress is possible in intelligence too, but it
eliminates the arguments against such a possibility.
Wishing to revolutionize the theory of intelligence, I found myself in
an urgent need for terminology and conceptual framework. Trying to make
some short-cuts, I adopted, as a starting point, the terminology and
concepts of philosophy of science. I found it suitable to start with because
of the apparent analogy between science and intelligence (which I
explained at length in my paper). I admit that it is not ideal for my purpose.
I even suggested some amendments. But still, I believe in its being a good
start more than a “hindrance”.
Thirdly, there is a message here for philosophy of science too. My
paper is, essentially, a case-study of a (relatively unstudied and
underdeveloped) special field of human activity – the field of intelligence. I
believe that philosophy of science can draw some lessons from this study
(in a way, intelligence is only one more way to study reality).
In sum, the detour that I took through philosophy of science had its
own reasons: (1) science sells good; (2) it pulls the carpet from under the
legs of some opponents of my call to reform intelligence theory; (3) it
supplies an initial stock of basic concepts and terms for that purpose; and
(4) it gives philosophy of science an opportunity to look inside a practical
field of (semi) scientific inquiry.
Having said all that, I must admit that I am fascinated by your idea
that the half-baked unacademic form of ‘criticism’ is more suitable for
intelligence than its philosophical mate. This idea seems very interesting to
me, and I feel that I need some time to ‘digest’ and explore it. Will it be too
much to ask you for some elaboration of this point? (for example: What is
the real character of the naive approach? i.e., what are the differences
between it and the scientific concept of ‘criticism’? How does it function
I also like very much your suggestion to replace intelligence officers
instead of trying to plant new ideas in their minds. But I wonder if it is
practical (well, maybe it will be so in a state of emergency, as was the state
of England during World-War-II).
Looking forward to hear from you,
and a happy new year,
yours sincerely,
Dr. Isaac Ben-Israel
Well, now that we seem to be involved in a longish debate, we should
omit inessentials and stick to basics, viz., first names, therefore,
22 January 1989
Dear Isaac,
I agree with you that science is not isolated but subjected to many
influences, philosophical influences among them. But while the
philosophical ideas that affected the sciences in the past were closely
connected with scientific practice and shared its fruitful imprecision, the
ideas that come from modern philosophy of science (up to and including
Popper and to some extent even Kuhn) are part of a school philosophy that
gives some general and very mislead outlines but never descends to details.
Curiously enough the trend was started by neopositivism which
prided itself of being a ‘scientific’ philosophy. However, ‘scientific’, for
neopositivists did not mean ‘in contact with scientific practice’ but ‘in
agreement with experience and the rules of logic’ where both ‘experience’
and ‘logic’ were defined in a very simplistic manner and independently of
scientific research. The debates between the various schools of
neopositivism (Quine and Popper included) may have been very dramatic
for those immediately affected – but they had no effect whatsoever on the
major scientific discoveries of the 20th century: relativity, quantum theory,
hadron-unification by quarks, the electroweak theory, the issue between
Big-Bang theories and Steady-State theories, the discovery of the structure
of DNA, the ‘New Synthesis’ in biology and on basic scientific debates
(such as the debate between Bohr and Einstein on the foundations of the
quantum theory). All they did was to give historically incorrect accounts of
the origin of relativity (here see the wonderful article by Holton on Einstein
and the Michelson experiment), of quantum theory and so to confuse
people instead of helping them. The only place where this mistaken and
simplistic philosophy is being taken seriously is in the ‘weaker’ subjects,
i.e., in some social sciences and here it is taken seriously only by people
who have no original ideas and think that methodology might help them
getting ideas. So, to summarize: philosophy is excellent if it is sufficiently
complex to fit in with scientific practice. The philosophies you mention are
not; they are “castles in the air” to quote Wittgenstein, they deceive people
but do not help them.
Are there philosophies of science I would accept? Yes, there are – and
they are being introduced by younger people who know science and its
history in detail and describe what is happening. Read for example Andrew
Pickering, Constructing Quarks and, especially, Peter Galison, How
Experiments End. Galison points out (a) that most existing philosophies of
science deal with theories and treat ‘facts’ or ‘experiments’ in a summary
way, (b) that experiments, especially large scale experiments such as those
carried out at CERN and other institutes (which are financed by
international agreements, involve hundreds of people and massive
equipment) have a life of their own and (c) that agreement concerning a
particular result and its ‘meaning’ is reached by a complex social process
whose features change from one experiment to the next. Item (c)) is very
interesting for your case for here, too, there are international agreements,
different groups are involved having different ideologies (in physics the
theoreticians think differently from the experimentalists and among the
latter the data evaluators think differently from data producers etc. etc.) and
the whole process is rather open ended. I would recommend to you to have
a look at Galison’s book and at an article by Holton, on Millikan and the
charge of the electron, mentioned in it. Pointing out that scientists, when
doing research, propose bold hypotheses and try to refute them is as
unenlightening in such cases as the remark that scientists, when doing
research, think – and as false: scientists often stick to timid hypotheses,
never mind the evidence and often proceed intuitively, without explicitly
discernible thought. I don’t think it is bad to provide them with rules of
thumb such as ‘try to falsify your hypotheses’ or ‘look for experimental
support’ which may be considered, but also disregarded, but it is deadly to
elevate such rules into ‘principles of rationality’ – but just that is being
done by the neopositivists and the Popperians.
Now, after this long speech (which, I hope, hasn’t exasperated you),
some details. On page 2 you say that4 “philosophy… helps a lot”. Well, my
first remark is that anything you consider in the sciences occasionally
helps, occasionally hinders research. This applies to mathematics,
experiment, philosophy and what have you. There are many episodes
where emphasis on experimental results impeded research and there are
other episodes where emphasis on mathematics led to empty talk (some
people believe that the so-called ‘theory of everything’ is such empty talk).
Same about philosophy, especially about school philosophies that were
constructed independently of scientific practice.
You say that you use the philosophy of science as a weapon against
two groups, those who confound facts with theory (group A) and those who
don’t believe that humans can conquer certain domains of reality (group
B). I would say that in using the philosophy of science you use a weapon
that is (a) unwieldy and (b) ineffective (except when you are dealing with
philosophers of science, of course).
4 See p. 11 above.
Lawyers at a trial convince a witness who says ‘but I saw it!’ that he
inferred, but did not see; they do this without any detour through three or
four different philosophies of science – and their arguments are much more
effective than any such detour would be. Why? Because they appeal to
commonsense and common experience and people can identify with it.
People, ordinary people – and I guess that many intelligence experts are
ordinary people in this sense – prefer appeals to commonsense to
theoretical shenanigans. A lawyer who brings up Popper just doesn’t know
how to conduct his case. As regards group B it suffices to mention an
example where a structure was found in an apparently very disorderly area
one example, a good and simple example achieves much more than even
the most sophisticated theoretical conversation. Philosophy of science as
practiced today simply is not good rhetoric (except for those already
immersed in it).
This is also the reason why I don’t regard the philosophy of science as
a good starting point for constructing a conceptual framework in a theory
of intelligence. First, because a theoretical framework may not be needed
(do I need a theoretical framework to get along with my neighbor?). Even a
domain that uses theories may not need a theoretical framework (in periods
of revolution theories are not used as frameworks but are broken into
pieces which are then arranged this way and that way until something
interesting seems to arise). And, secondly, because frameworks always put
undue constraints on any interesting activity. But I emphasize that this is
just a feeling of mine and there certainly does not exist any rule that forbids
you to start in this manner. I also feel that the inverse process is much more
promising: the philosophy of science certainly can learn a lot from what is
going on in hairy areas such as intelligence. As a matter of fact, I think that
the problematic nature of intelligence work gives us a much better idea
about man’s relation to ‘reality’ than physical science where things seem to
go much more smoothly.
Final page – 4 – you repeat my recommendation of a “half baked,
unacademic form of criticism”5 and say that you are fascinated by it. There
is another form of unacademic criticism that is not half baked but has a
long tradition behind it, and I mentioned it above: the criticism a clever
lawyer makes of the ‘evidence’ presented by a witness. I would strongly
recommend paying attention to this kind of criticism for it is much more
effective than the abstract considerations that emerge from the various
school philosophies of today. Concerning the kind of ‘evidence’ that comes
5 See p. 13 above.
up in intelligence matters a good lawyer is at least 100 times more effective
than a philosopher.
As regards my suggestion to replace intelligence officers by
intelligent people from other areas, actors, academics and so on, I
recommend to you the book A Man called Intrepid which describes the
situation in Great Britain in the Second World war. Noel Coward, Leslie
Howard, Turing were all involved and did much better than the intelligence
establishment which was tied down by tradition and silly rules. Of course,
it was a state of emergency but I think the same is true of the Near East of
Well, that is all for today. Let me conclude by saying that a case study
of particular intelligence episodes would be an excellent way of improving
the arid generalities of much of what goes for philosophy of science of
today, so you should really invert your program; not, what can the
philosophy of science do for intelligence, but what can intelligence do for
the philosophy of science.
All the best!
15 April 1989
Dear Paul,
First, I owe you an apology for taking so long to respond to your last
letter. I was away some months, and I am only now getting to read through
the mail. However, let me go directly to our debate (as you did in your last
On the General Situation
Let me begin with summarizing the general situation so far. In
response to my paper on “the Logic of Intelligence Process” (LIP), you
questioned the necessity of the detour I made through philosophy of
science. To this I gave 3 different answers: (1) I wanted to convince the
professional community to reform intelligence theory and thus I chose
science as a paradigm (because it sells well and supplies strong arguments
against certain groups of opponents); (2) philosophy of science provides
off-the-shelf theoretical framework ready for (almost) immediate
application to intelligence; and, finally, reversing the direction of my
arguments, (3) there is a lesson here for the philosophy of science to learn
from my case study in a remote field of ‘science’ (intelligence).
In response, you rejected reason (1), questioned (but not rejected) the
validity of (2) and accepted (3). According to your view, (a) science
doesn’t sell so well (lawyers sell a lot better); (b) theoretical frameworks
are not always needed and their necessity for intelligence is questionable.
Furthermore, philosophy of science (at least the Popperian school) is empty
and cannot do any good.
Well, one good reason is enough. In fact, I can accept your criticism
entirely and still justify my way (‘the detour’) because of (3) above.
Nevertheless, discovering some years ago that a complete agreement
is not a sufficient reason for stopping a good debate, I feel that trying to
counter you on your own ground will be beneficial for the subject I am
interested in. So, let me try.
On the Value of Popper’s Philosophy
As a physicist, I have great sympathy toward your statement about the
emptiness of what you call ‘modern school philosophy of science’
(including, and most especially, Popper’s). The emphasis it puts on logical
considerations, at the expense of details, is really disturbing (as a physicist,
it is even more disturbing to read papers of certain ‘philosophers’ about the
implications of, let say, quantum theory, when it is clear that these authors
do not have the slightest idea what it really looks like).
I agree that Popper’s philosophy is wrong. In fact, I rejected it in my
paper (LIP). But, you say much more: it is not only wrong, but misleading
and useless as well. Well, I admit that sometimes it is misleading; however,
it is not always the case (depending on the attitude of its audience). But, I
don’t think it is always useless. I am afraid that here you miss one of the
most prominent characteristics of Popper’s philosophy: it is always
stimulating. Sure it is wrong, and it takes time to show its falsity. But, the
process of exposing its fallacies is a fruitful one. My LIP, and even some of
your own best papers (e.g., “Consolations for the Specialist” or Chap. 9 in
your Philosophical Papers, Vol. 2) are good demonstrations of this.
On Emptiness and Informativeness
If I remember well, you once said that the content of philosophy of
science evaporated with the shift taken by logical-positivists (including
Popper) from details to structure of scientific theories (I don’t remember
the exact reference).
This line of reasoning seems to be a product of positivistic attitude by
itself: it is based on the famous positivistic Analytical Thesis dividing all
statements into two groups – empirical and formal (analytic, logical, you
name it). Taking this assumption for granted, it is clear that any philosophy
can be either informative (that is – based on empirical evidence) and
therefore, non-logical and perhaps even irrational, or logical, that is, non-
informative (empty). In fact, a teacher of mine in the past and a colleague
at present, Professor Zev Bechler, divides all possible philosophies into
these two types (he calls them Platonic and Aristotelian accordingly).
I think that these distinctions are based on wrong assumption (the
Analytical Thesis). Well, it is well known that Kant thought there is a third
group (the synthetic a priori), but he is not so popular nowadays, after the
collapse of some of his central examples (Newton’s physics, Euclid’s
geometry, etc.).
Personally, I don’t think he was totally wrong. I agree that his
examples are outdated, but not his basic ideas. In fact, I think that his
philosophy of science can be updated (and I even carried out part of this
program in my Ph.D. thesis). Kant knew something, which was forgotten
by the logical-positivists: he knew that philosophy of science cannot be
detached from philosophy of man. For him, philosophy of science,
epistemology and metaphysics meant the same. This is the reason that in
his Logic (edited by Jaesch in 1800) he adds to his famous three questions
(what can I know? what can I do? What can I hope for?) a forth one – what
is man? and he says that “the first question is answered by metaphysics, the
second by ethics, the third by religion and the fourth by anthropology.
Basically one could count all these to anthropology, because the first
three questions relate to the last one6. Unfortunately I have written on
this subject only in Hebrew. However, enclosed7 is a draft of an unfinished
paper on Kant’s philosophy of science as I interpret it. I would like very
much to hear your opinion on it.
If I am right, there is an option here for philosophy of science which
is relativistic (in the epistemological level) and yet realistic, and at the same
time linked tightly to philosophy of man.
On Scientists vs. Lawyers
Now, let me come back to your remarks.
I have already said that I like very much your idea about the practical
criticism of, let say, Columbus, sailing westward to ‘test his worldview
(‘theory’). I still feel that this idea should be explored thoroughly.
However, in your last letter you say that
“There is another form of unacademic criticism that is not
half baked but has a long tradition behind it […]: the
6 My underlining.
7 The paper, "Kantian Metaphysical Foundation of Relativity Theory" is not fully
given here, since it played no major role in the followings. Instead, I attache only the
annex to this draft, since it bears some relevance to topics discussed below.
criticism a clever lawyer makes of the ‘evidence’ presented
by a witness”,
and you
“strongly recommend paying attention to this kind of
criticism for it is much more effective than the abstract
considerations that emerge from the various school
philosophies of today. Concerning the kind of ‘evidence’
that comes up in intelligence matters a good lawyer is at
least 100 times more effective than a philosopher.”
Well, I am afraid that I don’t find this recommendation to be suitable
neither for intelligence nor for science. Let me explain why.
Lawyers, as you rightly notice, are experts of persuasion. But their
techniques ‘work’ only in a well-defined environment, namely, the court.
The ‘rules of the (legal) game’ are carefully formulated, known to every
‘player’, and that is essential for their work. Actually, they rarely convince
the jury that the defendant is not guilty; instead, they convince the jury that
they cannot find him guilty given certain laws. Can you imaging their work
without a detailed codex, trying to prove that their client is innocent
without knowing the rules of the game? The story of Joseph K is a fine
demonstration of this absurdity. Intelligence officers (and scientists as well)
don’t have any ‘rules’ that can be taken for granted. For them, everything
is up for grabs (at least it should be so).
Thus, I don’t think that the ‘criticism’ applied by lawyers is suitable
for our purpose: it is a sterile one, effective only when the ‘enemy’ and the
‘war’ rules are given beforehand. Scientists and intelligence officers do not
have this luxury. They have to struggle with unknown enemy (or Nature)
and there are no real constraints for this secret war. Anything goes!
On Intelligence in the Second World War
This character of intelligence work has been remarkably demonstrated
by the British secret war during the Second World War. You mentioned
(twice) the book A Man Called Intrepid, telling the story of Sir William
Stephenson (by the way, he passed away last month). I read the book many
years ago and I share your appreciation of it. Nevertheless, I think that the
best description of British intelligence during that period, from the point of
view of estimate and analysis, is given by R. V. Jones in his Most Secret
War. Jones was the man who invented and founded the disciplines of
scientific intelligence, starting the war as a junior officer and ending it at
the top of British intelligence as Director of Scientific Intelligence. Reading
carefully his story, it is evident that his ‘secret of success’ was the way he
used his head. He certainly used it differently (see chap. 37 for example).
In a way, his story supports your recommendation to replace old
cadres by young fresh minds; but I think it also proves this is not sufficient.
The new system may suffer from a lack of experience and be worse than
the older. Usually, what is needed is some combination of fresh minds
with a professional experienced establishment. Of course, the relative
weight of these two should be carefully balanced (one element can easily
overweight the other).
All in all, I do believe that the story of British intelligence during the
Second World War supports my view concerning the type of reform needed
in intelligence theory.
By the way, both Jones and Stephenson were recruited and introduced
to Churchill by Professor Frederick Alexander Lindemann (1886-1957),
who served together with Sir William Stephenson as pilots in the same
squadron in First World War, and who supervised at Oxford the Ph.D.
thesis in physics of young Mr. Jones between the wars.
Professor Lindemann, (later Lord Cherwell) is also the father of
another idea which is relevant to my subject: operations research, that is,
applying scientific methods to questions of war which were considered, up
to that time, as an art. The outstanding success of this technique should
teach us some profound lessons about the (non) demarcation of science and
Well, I feel that a new subject emerges here. So I will hold myself and
keep it for the next occasion.
Looking forward to hear from you,
yours sincerely,
Isaac Ben-Israel
Kantian Metaphysical Foundation of Relativity Theory (A Draft)
Isaac Ben Israel
Annex: Summary of Kant's Philosophy of Science
We can summarize Kant's philosophy of science with the following
(1) Let us consider a certain law of nature, L. L is either true or false. In
case it is true, then, being a law, it must also be necessary.
(2) The Copernican Assumption: If L is necessary (i.e. it is a true law of
nature), then its necessity can have no source other than the structure
of human cognitive mind.
(3) Therefore, we can use L to reveal this structure. We can use the
alledged necessity of L in order to derive a corresponding principle, P,
which governs the human cognition. Doing so, we should always bear
in mind that:
(a) P is 'derived' from L in the same way as theoretical
entities are 'derived from factual evidence. That is, P is merely a
hypothesis that enables us to explain certain phenomena of
(b) If L proves to be false, that is - if our belief in its
being a true law of nature turns to be not so well founded, then
we should have to abandon P as well.
(c) P will be a suitable and well-established hypothesis if
we succeed in showing that it is a necessary condition for L, i.e.
that L could not be a true law of nature unless P holds.
(4) Such a proposition P is synthetic a priori:
(a) It is synthetic because it is not logically or linguistic
necessary (we can perfectly imagine a consistent possible world
populated with creatures that have a totally different cognition
faculties from ours).
(b) Experience cannot confirm or falsify it, because experience itself
is based on P (as a result of the Copernican assumption).
Therefore P is a priori. In other words, P is epistemically
necessary, i.e. it is necessary for (and only for) the human mind
with its own peculiar understanding.
(5) The method, described above, of deriving synthetic a priori
principles from experience is called 'transcendental' method.
(6) Using the set {P1, P2, P3…} of synthetic a priori principles that
were obtained by the transcendental method (from the laws L1, L2…),
we can continue and logically derive from it some more propositions,
S1, S2, etc. All these propositions will be a priori, because they reflect
the principles that govern human understanding, and therefore their
truth does not depend on the specific content of our experience.
(7) We must distinguish carefully the epistemic status of the
principles P1, P2… from the status of their justification. Every Pi is a
priori although the propositions that take part in its justification are
based on experience.
(8) Therefore, No one of these principles (not P1, P2… nor S1, S2…)
is (absolutely) certain. Their truth is based on an empirical basis (i.e.
the truth of L1, L2…) and hence it can be doubted. The synthetic a
priori principles are hypotheses about our mind, which must be tested
against our experience. But of course, if their empirical basis is firm,
then they will be firm and certain as well.
(9) One can show that every Li is a priori in itself, because it can be
derived from the synthetic a priori principles P1, P2… This procedure
is, of course, circular. Nevertheless, it is not worthless. It completes
the analytical side of the derivation and reveals how the fundamental
laws of nature are consequences of our formal conditions of
understanding; and thus it demonstrates how a certain law of nature
(namely, the vast observation data it carries) can be counted as
supporting evidence in favor of the hypotheses P1, P2
30 May 1989
Dear Isaac,
While you were gallivanting about I got married (for the fourth time,
but for the first time seriously) and now my first priority is to find a home
for both of us (in Italy – my wife is an Italian), then to found a family and
then to resign from all my jobs to take care of it or, at least, to linger
nearby, in case some catastrophe arises. Still, I at once reply to part of your
letter (not to the Kant essay – that one I shall read later).
Starting from the end I most enthusiastically subscribe to the ‘new
subject’ which is going to emerge from the non-demarcation of the sciences
and the arts-humanities. As a matter of fact, my basic objection to any
philosophy of science that constructs a methodological system is that it
overlooks the art-aspect of science which requires rules of thumb, lots of
them rather than methodological principles.
[Popper is ambiguous on this point; on the one side he started his
lectures at the LSE in 1952 with the comment: “I find myself in a
paradoxical situation; I am a professor of scientific method – but there is no
scientific method; there are only rules of thumb”; and then he proceeded to
develop his falsificationism. On the other hand he often wrote and spoke
as if overruling this set of rules of thumb was a crime against reason
herself. I accept falsification as a rule of thumb. It is a very old rule: for the
sophists the use of counter examples was one of the most efficient ways of
advancing an argument. Ancient philosophers also knew about its
limitations; thus Plato, in various of his dialogues called the brute use of
counter examples antilogike – word bashing, and recommended a more
sophisticated procedure. It is a very useful rule – but it is not a condition of
rationality and’ historically, falsification is not the most frequent and most
efficient motor of scientific change (as Popper asserts in his Postscript).]
The art aspect of science becomes very clear from research done by a
new (post-Kuhnian) generation of historians. An example is Peter Galison,
How Experiments End (do you know the book? I found it most
interesting). Here he discusses three episodes, the Einstein-de Haas
experiment, the discovery of the muon, and the acceptance of neutral
currents. The important point is that the distinction between a context of
discovery and a context of justification, so important for principle-ridden
philosophies of science, simply does not exist and, that unanimity
concerning a certain effect is the result of developments and debates that
have much in common with what precedes the conclusion of a political
treaty; no party is really satisfied, there are compromises, because there are
many parties (the Western group under Millikan and an Eastern group in
the muon case)]. And a ‘fact’, then, is the result of such compromises on
the basis of shared or partly shared rules of thumb. Reading these stories
(and Andy Pickering’s Constructing Quarks) made it clear to me that a
principle-bound philosophy of science just finds no point of attack in the
scientific material. It is not ‘false’ – it is irrelevant.
But – and here I now accept your view – it is not therefore entirely
useless. A scientist may find comfort or inspiration in relating some of his
rules of thumb to a methodological system and in modifying them as a
result. What affects his research, still are rules of thumb; but what affects
his acceptance of some rules of thumb over others and what gives him the
confidence to use unusual rules of thumb may be a system. However,
applying the system directly, or turning the suggested rules into principles
because they come from the system – that is a very dangerous thing. This is
how I summarize, for my use, what you write in the bottom part of page 2.8
And this is the extent to which I grant a ‘stimulating’ quality to popper’s
philosophy, insofar as he regards it as a system.
I emphatically agree with the need to view epistemology etc. as parts
of anthropology (though I would rather say ‘politics’, for anthropology is a
special subject run by intellectuals while democratic politics, ideally, is run
by all). I always was very impressed by the way in which Aristotle
criticized Parmenides’ arguments in favor of an unmoving and indivisible
ONE. There were tow criticisms. The one was logical: it looked at the
argument and tried to show its faults. But there was a second criticism
which I would formulate as follows: Parmenides’ ONE makes nonsense of
life in the city hence, anybody who chooses life in the city has to reject
Parmenides. For me this means that the definition of what is real and
what is ‘mere appearance’ (or what is objective and what is subjective)
depends on what kind of life one wants to live. It is the result of a
political decision. Hence, epistemology without politics is incomplete and
arbitrary. Adding anthropology is a little better but still remains within the
domain of thought of a small minority.
Lawyers: I agree with your objection but add that lawyers can also
work outside well-defined frames and that their practice inside the frame
gives them an experience that is better suited for the discussions of
intelligence problems than the experience of philosophers of science which
is purely abstract and out of touch with human nature. (Joseph K is about
the system only.) I also agree that while fresh blood is good, a combination
of fresh blood and professional experience is even better provided the
8 See p.21 above.
institutional arrangements neither offend the experienced old-timers nor
give them too much power. But that is difficult to do.
Thank you for mentioning Jones’ book; I had heard of it, but now I
know a little more and shall try to get it. One thing is sure: the field of
intelligence certainly is an excellent testing ground of rules of thumb,
principles, methodological systems.
And now I have to run. All the best – I’ll soon write again. Long time
ago I had some exchange with Zev Bechler. I also used some of his papers
in my class on philosophy of science (mainly what he wrote about
Newton). Give him my regards!
15 June 1989
Dear Paul,
Let me open this rather short letter with an old Jewish greeting: Mazel
Tov (good luck) for you and your new wife [by the way, is her name
Grazia? The intelligence estimator inside me couldn’t resist the temptation
to guess her name…].9 Please give my heartiest congratulations to her.
Interesting enough, our ancient fathers probably thought that what one
needs for a successful marriage is luck (but I guess I don’t have to tell this
old wisdom to a man who gets married for the fourth time, even if the
previous three were not serious…).
Marriage, after all, is a serious business. So I shall not mix it with any
philosophical chatting and postpone my reply to your last letter to another
opportunity (maybe after you will comment on my unfinished paper on
Kant). Meanwhile I shall have time to finish reading Galison’s How
Experiments End (you see, I followed your recommendation and I’m
reading it now).
So, Mazel Tov again,
and looking forward to hear from you,
yours sincerely,
Isaac Ben-Israel
9 I added this note at the bottom of my letter. My (successful) guess of her name was
based on the acknowledgement of Feyerabend in his book, Farewell to Reason
(Verso, 1987 two years before the start of our correspondence) to “my beautiful,
good and very patient friend Grazia Borrini”.
5 October 1990
Dear Paul,
It has been a long time since I received a letter from you. Since my
last one included only greetings for your marriage, I decided not to wait
any more for a reply (and the promised critique of my paper on Kant) and
to comment on your last letter from May 1989.
By the way, my paper on “the logic of estimate Process” (LEP), which
started our exchange of letters, has meanwhile been published, and a
preprint is enclosed. Unfortunately it was too late to incorporate some of
your comments but I managed however to make some minor changes (cf.
Note 28 in p. 715).10
Your remark on the way Aristotle criticized Parmenides’ ONE is very
interesting and raises immediately a host of important questions (by the
way, can you give me the exact reference of what you call Aristotle’s
‘political argument’? unfortunately I couldn’t locate it). Let me write down
few of these questions.
First, and perhaps the most important, - what exactly do you mean by
saying that “what is real and what is ‘mere appearance’ (or what is
objective and what is subjective) depends on what kind of life one
wants to live”? How does one decide what is real? Can he decide anything
he wants (provided it fits his needs), or is he limited by some external
10 This note reads as follows: “The critical method was studied in philosophy of
science, but it was not created there: ‘it is old hat; it was used by the discoverers of
new continents, by businessmen like Marco Polo, by Generals like Clausewitz […]
and it was matter of course for the native tribes in Kenya (in the 1930s) who, being
faced with invaders of the most varied kind became more critical than the local
missionaries who met only other missionaries. […] This ‘naively’ (i.e. unacademic)
critical approach will most likely be more effective for it can also work in partly
closed surroundings. Businessmen were and still are rivals, there can be a flow of
information in some direction, not in others and so criticism here is adapted to
conditions necessary for intelligence work but absent (to some extent!) from the
sciences’ (Paul Feyerabend, a private letter to the author, 30 November 1988).”
constraints? In other words, is the definition of what is real arbitrary or is
there a supreme judge, namely, the external world itself?
Well, one can immediately reply, as you do (following Protagoras) in
Farewell to Reason (FtR, p.44), that although every society can choose
and define its own ‘reality’ in accordance with its own specific needs, it
doesn’t mean that this choice is arbitrary: after all, it should fit those needs!
Relativists, like Herodotus and Protagoras, don’t have to assert “that
institutions and laws that are valid in some societies and not valid in others
are therefore arbitrary and can be changed at will […]. One can be a
relativist and yet defend and enforce laws and institutions” (ibid.).
Still, this doesn’t answer the question of external factors. The
‘reality’ (or to use a more appropriate term, the ‘objectivity’) of social laws
and institutions is one thing, and the reality (and objectivity) of physical
phenomena is another thing. One can easily be a relativist in social and
cultural matters and, at the same time, be a ‘metaphysical’ realist (that is,
believe that the laws and ontology of the external world, unlike social laws
and institutions, do not depend on the way we want to live).
You argue (quite convincingly, allow me to say) for social and
cultural relativism: your relativism is “about human relations” (FtR, p. 83).
This is a popular doctrine today (as you notice yourself, cf. FtR, p. 77).
You also argue for epistemic relativism. This quite convincing too; after all,
how can we be sure that we know the truth, even if there is one? The
question I would like to formulate now is whether this contradicts
metaphysical (ontological) realism (concerning the physical world). In
other words, does social or epistemic relativism necessarily imply
metaphysical (i.e. ontological) relativism?
I admit that I don’t see any necessary logical relation between the tow:
it seems clear to me that arguments for social, cultural and even epistemic
relativism do not establish the case against (scientific) realism. We can
easily decide what social laws and institutions we would like to have, but
we cannot do it, at least to the same extent, with physical laws. As I see it,
in the case of external world you cannot (‘politically’) decide what there is.
The external world confines us to a very narrow space of liberty. True,
there is latitude here. Reality doesn’t force us to hold one and only one
view about it; but this latitude is not unlimited. One cannot arbitrarily
decide whatever one likes.
Take for example Galison’s How Experiment End. You mention it in
your letters and papers (cf. “Realism and the Historicity of knowledge”,
The Journal of Philosophy, Vol. LXXXVI, No. 8, August 1989)
recommending it as worth reading and as a good example of new and better
philosophy of science. I enjoyed reading it although I had already known its
moral – that experimental ‘facts’ are sometimes the results of debates and
compromises between rival groups on the basis of partly shared beliefs
(ibid., p. 394. I guess any physicist who actually practiced experimental
physics knows it). This fading of the line between the context of discovery
and the context of justification, exactly in the sense Galison describes, is a
direct result, I think, of the theory-ladenness of ‘observations’.
In fact, I have already written in my LEP (pp. 672-673) that:
“Any observation report, including those in physics, is
theory-laden; it can never be ‘pure’ observation and always
presumes certain hypotheses. Every observation contains, apart
from sensual element, an element of interpretation. […]
This problem of interpretation is common to many kinds of
human research and knowledge. It is not more acute in
intelligence than in the physics of elementary particles, for
example, where no one can ‘see’ (with the naked eye) any
particles and where any observation depends on clusters of
theories, hypotheses and interpretations of measurements”.
Galison makes an extraordinary (and successful) effort to show how
strong is the mutual interaction between theoretical presuppositions and
actual ‘results’ of experiments. One can read his study as a historical
account confirming the views of Duhem, Kuhn and you. However, I
believe that the main lesson one can derive from it goes, perhaps, against
the dominant line of contemporary philosophy of science. True,
‘experimental results’ are a product of negotiations between rival parties,
but only to a certain extent. It is very interesting to notice that Einstein
‘measured’ the very g=1 (in Einstein-de Haas experiment), he theoretically
expected, instead of the now believed value of g2. But, what is more
interesting is that even such an authority as Einstein’s (in 1915!) was not
enough for the scientific community to stop experimenting and publishing
different results. Doesn’t it show that one cannot arbitrarily decide
whatever he likes?
Reading your papers through the years, I always had the feeling that
you tried to stick to (ontological) realism despite your preaching for
epistemic relativism (sometimes you even expressed this tendency
explicitly; Cf. Chaps. 2 &11 in your Philosophical Papers, Vol. 1, [PP1]).
I must admit that after reading your FtR, I’m not so sure about this any
more. Perhaps you will care to comment on this issue.
Before going on to other matters, allow me to describe my view on the
question of realism-relativism. Basically I am an empiricist (who isn’t?),
that is, I believe that everything we know about the external world comes
through our senses (filtered, processed and distorted as it is by our
‘internal’ cognitive faculties). Every (scientific and non-scientific)
knowledge we have depends on some (partly explicit and mostly implicit)
assumptions and is true only as long as we accept some other propositions
as true. Knowledge of the external world is always relative. Now, if so,
what is the point in holding a realistic position? If it is clear that we cannot
know the external world ‘as it is’ (to use Kant’s phrase), why assume that
there is a ‘world’ which is causally independent of us, the observers?
In order to start answering this question one should realize that
everything we know is only a conjecture: If it is true that the world is
going in that way and not in this, then a certain proposition P is supposedly
true. There is no ultimate established body of knowledge. No knowledge
can be more than a theory (i.e., a set of conjectures) about the world, and
this includes philosophical contentions as well.
Now, how do we judge ‘theories’? What are the criteria for holding
this theory and not that? There are many possible replies to this question. I
prefer the one which judges a theory by its compatibility to some basic set
of (accepted) ‘facts’ (if it doesn’t, there is a problem somewhere in the
theory or in the set of ‘facts’ or in both).
So, why do I hold epistemic relativism? Because it is the only position
(theory) I know which fits certain basic ‘facts’ (historical facts, logical
arguments, etc. Your Against Method [AM] is full of such ‘facts’). I other
words, I hold epistemic relativism because it is a plausible theory. It
explains a lot of ‘facts’ I ‘know’.
And do I want to hold metaphysical realism? Exactly for the same
reason. It is too a (very) plausible theory without which it would be
difficult to understand certain ‘facts’; for example – the universal and
overwhelming agreement between human beings as to certain simple
observations of what happens in the world (“The conception of a world that
really exists is based on there being far-reaching common experience of
many individuals, in act of all individuals who come into the same or a
similar situation with respect to the object concerned” wrote Schrödinger to
Einstein in 18 November 1950).
Another strong ‘evidence’ for realism is the existence of error.
Without realism it would be highly difficult to explain the ‘fact’ that we do
err in our judgments of the external world.
Summarizing the arguments above, there are very good reasons for
being a realist and relativist simultaneously. Furthermore, I believe that
it is even possible. In fact, my main task in my Ph.D. dissertation was to
construct a multi-layer model of modal propositions (I call it ‘the Modal
Onion Model’ – MOM) and use it to demonstrate the possibility of holding
a certain proposition to be necessary and contingent at the same time (of
course not for the same type of necessity). The model enables one to be a
relativist and realist at the same time coherently. I would like to elaborate
on it a little, but before doing so, let me discuss first the relationship
between realism and relativism.
Following your general advice, let me discuss this relationship
through a concrete example: motion. Newton’s concept of motion is
absolute and real. Einstein’s concept is relative, yet as real as Newton’s.
Hence, at least in the case of motion, ‘relative’ is not the opposite of
‘real’ but of ‘absolute’. I think that this conclusion can be generalized:
relativism and absolutism are exclusive; relativism and realism are not.
Now, you condemn absolutism. Science, you claim, “never obeys, and
cannot be made to obey, stable and research independent standards” (PP1,
p. xiii), that is, there are no absolute standards in science. As a matter of
fact, there are no absolute standards in any field of human activity. In your
FtR you write: “The assumption that there exist universally valid and
binding standards of knowledge and action is a special case of a belief
whose influence extends far beyond the domain of intellectual debate. This
belief may be formulated by saying that there exists a right way of living
and that the world must be made to accept it” (pp. 10-11), and you observe
that this belief was the driving force behind many evils which were done
through the history of human kind.
So far so good. Absolutism has to be rejected. Hence, relativism has
to be preferred. But, doing so, why do we have to reject realism? Your
claim that “realism […] reflects the wish of certain groups to have their
ideas accepted as the foundations of an entire civilization and even of life
itself” (PP1, p. xiii) confuses between realism and absolutism. Surely one
can be a realist (i.e. believe that the world exists independently of our
knowledge of it) and still admit that this knowledge is fallible and not
I planned to go on and elaborate here on my Modal Onion Model, but
I realize now that this letter became rather long. So I will keep the rest of it
for the next exchange of letters.
Mazel Tov again,
and looking forward to hear from you,
yours sincerely,
Isaac Ben-Israel
11 October 1990
Dear Isaac,
I am about to leave for a trip to the South of Switzerland, California
and Italy, therefore I cannot give you a detailed reply to your letter.
However I enclose a paper11 that deals with precisely the problems you
raise and may, perhaps, give an answer here and there.
In your terminology – I do indeed assert that social relativism entails
metaphysical realism, though only to some extent. [The underlined
passage describes the extent to which I now differ from that said before:
relativism is possible because the world permits it – to a certain
extent]12. My argument is a metaphysical argument: reality (or Being) has
no well-defined structure but reacts in different ways to different
approaches. Being approached over decades, by experiment of ever
increasing complexity it produces elementary particles; being approached
in a more ‘spiritual’ way, it produces gods. Some approaches lead to
nothing and collapse. So I would say that different societies and different
epistemologies may uncover different sides of the world, provided Being
(which has more sides than one) reacts appropriately. I know, all this sound
quite mystical but I think it can be worked out to sound more plausible. At
any rate, the typescript13 and the printed paper are first steps.
Incidentally, there is no way of finding out the limit to which the
world permits relativism because Being itself cannot be known (I have
argument for that, too). What can be known is manifest Being, i.e. the
response of Being to a particular approach.
I looked briefly into your printed paper which looks very good. I
noted you ascribe proliferation and criticism based on it to Popper. This is a
little unjust to Mill (On Liberty) who gave much better argument for both.
Best wishes!
11 The paper enclosed was P. Feyerabend “Realism and the Historicity of Knowledge”,
The Journal of Philosophy, Vol. LXXXVI, No. 8, August 1989.
12 Feyerabend added this note at the bottom of his letter.
13 The typescript that was enclosed had the title “Ethics as a Measure of Scientific
Truth” and was dated from 16 August 1990.
25 October 1990
Dear Isaac,
You ask what I mean when saying that what is real and what is not
depends on the kind of life one wants to lead. Homer does not have a single
great distinction real/apparent. His world contains many different things,
people, animals, thunderstorms, gods, dreams among them. They all can be
experienced, though under different and sometimes very complex
conditions (just think how difficult it is to get a glimpse of a very shy bird -
–you can hear him, but to get a full view of him is almost impossible). Now
the Presocratics and then Plato replaced this complex net by a bifurcation
Why? In Plato the reason is clear – it is political: to have the stable
society which, according to him, is needed for happy human beings, you
have to tie your existence to what is stable. These are the most important
ingredients and Plato plus some Presocratics call them by names, which
one could roughly translate as ‘real’.
Much more recently some molecular biologists, Delbruck among
them, have suggested to abolish botany, zoology etc. in favor of a
comprehensive molecular biology. Now persons for which direct contact
with people, plants, and animals forms the basis of their lives will not reject
molecular biology. They will regard it as adding interesting details and will
interpret it as an instrument of prediction, when it conflicts with thw macro-
phenomenological way of life.
And so on: the ‘external world’ is that part of the many strange
phenomena that surround us which conforms most closely with our most
important beliefs, attitudes etc.
Think of ‘health’ as an example. For some, ‘health’ is a good working
condition of the body-machine – for others it is happiness and benefit for
others – even if the body is not quite what it ought to be according to the
Now I don’t think this is relativism. Relativism presupposes a fixed
framework. But people with different ways of life and different conceptions
of reality can learn to communicate with each other, often even without a
gestalt-switch, which means, as far as I am concerned, that the concepts
they use and the perceptions they have are not nailed down but are
ambiguous. They are ambiguous to an extent that you cannot even speak of
theory-ladennes – which assumes that there is something non-theoretical
carrying the load. The best thing is to drop this dichotomy
theory/observation which is still an aftereffect of the old belief that
something called ’observation’ is the final arbiter of everything. Where is
‘observation’ in Machamer’s more complex examples? Computers do the
dirty work and the scientist simple reads the results – he ‘observes’
computer printouts (see the enclosed short note14).
Of course, one cannot arbitrarily decide whatever one likes – at least
as long as one is part of a community. On the other hand, this community
has neither a wise man nor a wise process (‘experience’, or ‘experiment’) to
turn to, all it has are the results of temporary political treaties – and these
results will last only as long as the political situation remains fairly stable
(compare new experimental equipment with a new political ideology, like
You are right about my tendency to emphasize realism but I have
changed on that. I now distinguish between an ultimate reality, or Being.
Being cannot be known, ever (I have arguments for that). What we do
know are the various manifest realities, like the world of the Greek gods,
modern cosmology etc. These are the results of an interaction between
Being and one of its relatively independent parts (a section of history, a
tradition, an ambitious group like the group around Delbruck that started
molecular biology, or even an individual): what we claim to recognize id
not independent of us; what is independent of us is and remains
This is kind of an ontological relativism, which is possible because
Being is built accordingly: it reacts to some approaches, not to all, but not
to others. In this sense (which I have still to work out) I would say that
relativism plus realism are compatible.
How does that sound? And now I have to get back to my taxes.
Best wishes!
14 Enclosed was “Science without Experience”, taken from P. Feyerabend, Realism,
Rationalism & Scientific Method – Philosophical Papers Vol. 1, Cambridge
University Press, 1981, Chap. 7.
... Vgl. auch den Brief Feyerabends an Ben-Israel (2001: 98) vom 11.10.1990(Ben-Israel 2001; Hervorhebung im Original): "My argument is a metaphysical argument: reality (or Being) has no welldefined structure but reacts in different ways to different approaches. ...
... uenzen dieser Entscheidung tragen möchte -was zumindest bedeutet, dass es Nutznießer und Benachteiligte geben muss. Man ist im gleichen Atemzug verletzt wird; dem "Being" wird eine kausale Wirkungsmächtigkeit zugeschrieben "Nature responds" -noch dazu in mannigfacher Variation. Vgl. auch den Brief Feyerabends an Ben-Israel (2001: 98) vom 11.10.1990(Ben-Israel 2001; Hervorhebung im Original): "My argument is a metaphysical argument: reality (or Being) has no welldefined structure but reacts in different ways to different approaches. Being approached over decades, by experiment of ever increasing complexity it produces elementary particles; being approached in a more ‚spiritual' way, it produced go ...
Full-text available
Translation-both multi-and intra-lingual-is vital to anthropological method. Drawing a distinction between two opposing modes of translation ("domesticating" versus "foreignizing"), this paper considers the ontological and ethical consequences of these two interpretative strategies, in particular by critically engaging with the doctrine of Donald Davidson, the theoretical inspiration for João Pina-Cabral's work, World. I argue, instead, in favor of a "pagan" or pluralizing conceptual method, inspired by Feyerabend, Lyotard, and Hans Peter Duerr, and I suggest that their approaches demonstrate that even the polymodal on-tology of Latour is lacking in plurality. In conclusion, I consider how the notion of foreignizing translation relates to the method associated with the ontological turn in anthropology.
Full-text available
Artykuł bada ważniejsze argumenty przeciwko naturalizmowi metodologicznemu. Argumenty te mają stanowić podstawę do uzasadnienia następujących tez. Naturalizm metodologiczny: źle wpływa na rozwój wiedzy; utrudnia współzawodnictwo w nauce; jest tylko częścią określonej tradycji, która została zabsolutyzowana; jest wyłącznie prowizoryczną zasadą; jest arbitralną i szkodliwą regułą; jest podejściem irracjonalnym; jest złą filozofią; jest ujęciem przyjmowanym bezkrytycznie. Większość antynaturalistycznych argumentów nie jest przekonująca. Jednak kilka z nich można uznać za zasadne.
The goal of this paper is to provide an interpretation of Feyerabend's metaphysics of science as found in late works like Conquest of Abundance and Tyranny of Science. Feyerabend's late metaphysics consists of an attempt to criticize and provide a systematic alternative to traditional scientific realism, a package of views he sometimes referred to as “scientific materialism.” Scientific materialism is objectionable not only on metaphysical grounds, nor because it provides a poor ground for understanding science, but because it implies problematic claims about the epistemic and cultural authority of science, claims incompatible with situating science properly in democratic societies. I show how Feyerabend's metaphysical view, which I call “the abundant world” or “abundant realism,” constitute a sophisticated and challenging form of ontological pluralism that makes interesting connections with contemporary philosophy of science and issues of the political and policy role of science in a democratic society.
Feyerabend's interests in religion and mysticism grew through his career. In his later writings, Feyerabend's numerous critiques of scientific materialism are often accompanied by purported advantages of religious orientations and temperaments. These recommendations do not simply follow from his tolerant theoretical pluralism; they are more positive attempts to articulate distinctive aspects of human life satisfied by religion, but not by scientific materialism. Elevating the human need for mystery, reverence, and love, he contrasts these goods with the deliverances of monistic conceptions of science and reason. I bring attention to some of the common themes in these remarks to argue that they were integral with other parts of his philosophical project and that they could serve as helpful rejoinders to contemporary exhortations to science-based secularism from philosophers of science.
This paper explores the account of ‘ultimate reality’ developed in the later philosophy of Paul Feyerabend. The paper has five main parts, this introduction being the first. Part two surveys Feyerabend’s later work, locates it relative to his more familiar earlier work in the philosophy of science, and identifies the motivations informing his interest in ‘ultimate reality’. Part three offers an account of Feyerabend’s later metaphysics, focusing on the account given in his final book, Conquest of Abundance. Part four then assesses Feyerabend’s related claims that ‘ultimate reality’—or ‘Being’—is both ‘ineffable’ and ‘abundant’, and tries to reconcile these two ‘doctrines’ with one another. I conclude in part five that into his later period Feyerabend offers a positive account of ‘ultimate reality’ which identifies it as receptive to a radical plurality of modes of inquiry and forms of knowledge. Such ‘abundance’ arises from the interaction of human cognitive and creative capacities with ‘ineffable Being’ on the other, such that ‘ultimate reality’ itself remains ‘forever unknowable’.
This paper explores the influence of the fifth-century Christian Neoplatonist Pseudo-Dionysius the Areopagite (Denys) on the twentieth-century philosopher of science Paul Feyerabend. I argue that the later Feyerabend took from Denys a metaphysical claim—the ‘doctrine of ineffability’—intended to support epistemic pluralism. The paper has five parts. Part one introduces Denys and Feyerabend’s common epistemological concern to deny the possibility of human knowledge of ultimate reality. Part two examines Denys’ arguments for the ‘ineffability’ of God as presented in On the Divine Names. Part three then explores how Feyerabend imported Denys’ account of divine ineffability into his own metaphysics to provide a novel argument for epistemic pluralism. Part four explains the significance of an appreciation of Dionyius’ influence for our understanding of Feyerabend. I conclude that Denys was a significant and neglected influence upon the later Feyerabend.
Full-text available
Intelligence estimates based on models keyed to frequency and recency of past occurrences make people less secure even if they predict most harmful events. The U.S. presidential commission on WMDs, the 9/11 commission, and Spain's comisión 11-M have condemned the status quo mentality of the intelligence community, which they see as being preoccupied with today's “current operations” and tactical requirements, and inattentive to tomorrow's far-ranging problems and strategic solutions. But the overriding emphasis in these commissions' recommendations is on further vertically integrating intelligence collection, analysis, and operations. Such proposals to further centralize intelligence and unify command and control are not promising given recent transformations in Jihadist networks to a somewhat “leaderless resistance” in the wake of Al Qaeda's operational demise. To defeat terrorist networks requires grasping novel relations between an englobing messianic moral framework, the rootless intellectual and physical mobility of immigrant and diaspora communities, and the overarching conceptual, emotional, and logistical affordances of the Internet. Britain's WWII experience provides salutary lessons for thinking creatively with decentralized expertise and partially autonomous approaches.