Quantum Diaries

In this, the epilogue to my philosophic musing, I locate my view of the scientific method within the landscape of various philosophical traditions and also tie it into my current interest of project management. As strange as it may seem, this triumvirate of the scientific method, philosophy and management meet in the philosophic tradition known as pragmatism and in the work of W. Edwards Deming (1900 – 1993), a scientist and management guru who was strongly influenced by the pragmatic philosopher C.I. Lewis (1883 – 1964), who in turn strongly influenced business practices. And I do mean strongly in both cases. The thesis of this essay is that Lewis, the pragmatic philosopher, has had influence in two directions: in business practice and in the philosophy of science. Surprisingly, my views on the scientific method are very much in this pragmatic tradition and not crackpot.

The pragmatic movement was started by Charles S. Peirce (1839 – 1914) and further developed by Williams James (1842 – 1910) and John Dewey (1859 – 1952). The basic idea of philosophic pragmatism is given by Peirce in his pragmatic maxim as: “To ascertain the meaning of an intellectual conception one should consider what practical consequences might result from the truth of that conception—and the sum of these consequences constitute the entire meaning of the conception.” Another aspect of the pragmatic approach to philosophic questions was that the scientific method was taken as given with no need for justification from the outside, i.e. the scientific method was used as the definition of knowledge.
How does this differ from the workaday approach to defining knowledge? Traditionally, going back at least to Plato (428/427 or 424/423 BCE – 348/347 BCE) knowledge has been defined as:
1) Knowledge – justified true belief
The leaves open the question of how belief is justified and since no justification is ever 100% certain, we can never be sure the belief is true. That is a definite problem. No wonder the philosophic community has spent two and a half millennia in fruitless efforts to make sense of it.

A second definition of knowledge predates this and is associated with Protagoras (c. 490 B.C. – c. 420 B.C.) and the sophists:
2) Knowledge – what you can convince people is true
Essentially, the argument is that since we cannot know that a belief is true with 100% certainty; what is important is what we can convince people of. This same basic idea shows up in the work of modern philosophers of science with the idea that scientific belief is basically a social phenomenon and what is important is what the community convinces itself is true. This was part of Thomas Kuhn’s (1922 – 1996) thesis.

While we cannot know what is true, we can know what is useful. Following the lead of scientists, the pragmatists effectively defined knowledge as:
3) Knowledge – information that helps you predict and modify the future
If we take predicting and modifying the future as the practical consequence of information, this definition of knowledge is consistent with the pragmatic maxim. The standard model of particle physics is not knowledge by the strict application of definition 1) since it is not completely true; however it is knowledge by definition 3 since it helps us predict and modify the future. The scientific method is built on definition 3). The modify clause is included in the definition since the pragmatists insisted on that aspect of knowledge. For example, C.I. Lewis said that without the ability to act there is no knowledge.

The third definition of knowledge given above does not correspond to what many people think of as knowledge so Dewy suggested using the term “warranted assertions” rather than knowledge: The validity of the standard model is a warranted assertion. Fortunately, this terminology never caught on. In contrast, James’s pragmatic idea of “truth’s cash value”, derided at the time, has caught on. In a recent book “How to Measure Anything,” on risk management, Douglas W. Hubbard spends a lot of space on what is essentially the cash value of information. In business, that is what is important. The pragmatists were, perhaps, just a bit ahead of their time. Hubbard, whether he knows it or not, is a pragmatist.
Dewey coined the term “instrumentalism” to describe the pragmatic approach. An idea or a belief is like a hand, an instrument for coping. A belief has no more metaphysical status than a fork. When your fork proves inadequate to the task of eating soup, it makes little sense to argue about whether there is something inherent in the nature of forks or something inherent in the nature of soup that accounts for the failure. You just reach for a spoon . However, most pragmatists did not consider themselves to be instrumentalists but rather used the pragmatic definition of knowledge to define what is meant by real.

Now I turn to C.I. Lewis. He is alternately regarded as the last of the classical pragmatists or the first of the neo-pragmatists. He was quite influential in his day as a professor at Harvard from 1920 to his retirement in 1953. In particular, his 1929 book “Mind and the World Order” had a big influence on epistemology and surprisingly on ISO management standards. One can see a lot of the ideas developed by Kuhn already present in the work of C.I. Lewis , for example, the role of theory in interpreting observation. Or as Deming, influenced by Lewis, expressed it: “There is no knowledge without theory.” As a theorist, I like that. At the time, this was quite radical. The logical positivists took the opposite tack and tried to eliminate theory from their epistemology. Lewis and Kuhn argued this was impossible. The idea that theory was necessary for knowledge was not new to Lewis but is also present in the works of Henri Poincaré (1854 – 1912) who was duly reference by Lewis.

Another person Lewis influenced was Willard V. O. Quine (1908 – 2000), although Quine and Lewis did not agree. Quine is perhaps best known outside the realm of pure philosophy for the Duhem-Quine thesis, namely that it is impossible to test a scientific hypothesis in isolation because an empirical test of the hypothesis requires one or more background assumptions. This was the death knell of any naïve interpretation of Sir Karl Popper’s (1902 –1994) idea that science is based on falsification. But Quine’s main opponents were the logical positivists. Popper was just collateral damage. Quine published a landmark paper in 1951: “Two Dogmas of Empiricism”. I would regard this paper as the high point in the discussion of the scientific method by a philosopher and it reasonably readable (unlike Lewis’s “The Mind and the World Order”). Beside the Duhem-Quine thesis, the other radical idea is that observation underdetermines scientific models and that simplicity and conservatism are necessary to fill the gap. This idea also goes back to Poincaré and his idea of conventionalism – much of what is regarded as fact is only convention.

To a large extent my ideas match well with the ideas in “Two Dogmas of Empiricism.” Quine summarizes it nicely as: “The totality of our so-called knowledge or beliefs, from the most casual matters of geography and history to the profoundest laws of atomic physics or even of pure mathematics and logic, is a man-made fabric which impinges on experience only along the edges.” and “The edge of the system must be kept squared with experience; the rest, with all its elaborate myths or fictions, has as its objective the simplicity of laws.” Amen.

Unfortunately, after the two dogmas of empiricism were brought to light, the philosophy of science regressed. In a recent discussion of simplicity in science I came across, there was neither a single mention of Quine’s work nor his correct identification of the role of simplicity – to relieve the under determination of models by observation. Philosophers found no use for his ideas and have gone back to definition 1) of knowledge. Sad

Where philosophers have dropped the ball it was picked by people in, of all places management. Two people influenced by Lewis were Walter A. Shewhart (1891 – 1967) and Edwards Deming. It is said that Shewhart read Lewis’s book fourteen times and Deming read it nine times. Considering how difficult that book is, it probably required that many readings just to comprehend it. Shewhart is regarded as the father of statistical process control, a key aspect of quality control. He also invented the control chart, a key component of statistical process control. Shewhart’s 1939 book “Statistical Method from the viewpoint of Quality Control” is a classic in the field but it devoted a large part to showing how his ideas are consistent with Lewis’s epistemology. In this book, Shewhart introduced the Shewhart cycle, which was modified by Deming (and sometimes called the Deming cycle). Under its current name Do-Plan-Check-Act (DPCA cycle) it forms the basis of the ISO management standards.

The original Shewhart cycle as given in Shewhart’s book.

What is this cycle? Here it is as captured from Shewhart’s book. This is the first place where production is seen as part of a cycle and in the included caption Shewhart explicitly relates it to the scientific method as given by Lewis. Deming added another step to the cycle, which strikes me as unnecessary; the act step. It can easily be incorporated in the specification or plan stage (as it is in Shewhart’s diagram). But Deming was influenced by Lewis who regarded knowledge without the possibility of acting as impossible, hence the act step. This idea has become ingrained in ISO management standards as the slogan “continual improvement” (Clause 10 in the standards). To see the extent Deming was guided by Lewis’s ideas just look at Deming’s 1993 book “The New Economics.” He summarizes his approach in what he calls a system of profound knowledge. This has four parts: knowledge of system, knowledge of variation, theory of knowledge and knowledge of physiology. The one that seems out of place is the third; why include theory of knowledge? Deming believed that this was necessary for running a company and he explicitly refers to Lewis’s 1929 book. Making the reading of Lewis’s book mandatory for business managers would certainly have the desirable effect of cutting down the number of managers. To be fair to Deming, he does suggest starting in about the middle of the book. We have two unbroken chain – 1) Peirce, Lewis, Shewhart, Deming, ISO management standards and 2) Pierce, Lewis, Quine, my philosophical musings . It reminds one of James Burke’s TV program “Connections”.

Popper may be the person many scientists think of to justify how they work but Quine would probably be better and Quine’s teacher, C.I. Lewis, through Deming, has provided the philosophic foundation for business management. Within the context of definition 3) for knowledge both science and business have been very successful. Your reading of this essay required both. In contradistinction, standard western philosophy based on definition 1) has largely failed; philosophers still do not know how to acquire knowledge. However, not all philosophy is useless, some of it is pragmatic.

Isaac Asimov (1920 – 1992) “expressed a certain gladness at living in a century in which we finally got the basis of the universe straight”. Albert Einstein (1870 – 1955) claimed: “The most incomprehensible thing about the world is that it is comprehensible”. Indeed there is general consensus in science that not only is the universe comprehensible but is it mostly well described by our current models. However, Daniel Kahneman counters: “Our comforting conviction that the world makes sense rests on a secure foundation: our almost unlimited ability to ignore our ignorance”.

Well, that puts a rather different perspective on Asimov’s and Einstein’s claims.  So who is this person that is raining on our parade? Kahneman is a psychologist who won the 2002 Nobel Prize in economics for his development of prospect theory. A century ago everyone quoted Sigmund Freud (1856 – 1939) to show how modern they were. Today, Kahneman seems to have assumed that role.[1]

Kahneman’s Nobel Prize winning prospect theory, developed with Amos Tversky (1937 –1996), replaced expected utility theory. The latter assumed that people made economic choices based on the expected utility of the results, that is they would behave rationally. In contrast, Kahneman and company have shown that people are irrational in well-defined and predictable ways. For example, it is understood that the phrasing of a question can (irrationally) change how people answer, even if the meaning of the question is the same.

Kahneman’s book, Thinking, Fast and Slow, really should be required reading for everyone. It explains a lot of what goes on (gives the illusion of comprehension?) and provides practical tips for thinking rationally. For example, when I was on a visit in China, the merchants would hand me a calculator to type in what I would pay for a given item. Their response to the number I typed in was always the same: You’re joking, right?  Kahneman would explain that they were trying to remove the anchor set by the first number entered in the calculator. Anchoring is a common aspect of how we think.

Since, as Kahneman argues, we are inherently irrational one has to wonder about the general validity of the philosophic approach to knowledge; an approach based largely on rational argument. Science overcomes our inherent irrationality by constraining our rational arguments by frequent, independently-repeated observations.  Much as with project management, we tend to be irrationally overconfident of our ability to estimate resource requirements.  Estimates of project resource requirements not constrained by real world observations leads to the project being over budget and delivered past deadlines. Even Kahneman was not immune to this trap of being overly optimistic.

Kahneman’s cynicism has been echoed by others. For example, H.L. Mencken (1880 –1956) said:  “The most common of all follies is to believe passionately in the palpably not true. It is the chief occupation of mankind”. Are the cynics correct? Is our belief that the universe is comprehensible, and indeed mostly understood, a mirage based on our unlimited ability to ignore our ignorance? A brief look at history would tend to support that claim.  Surely the Buddha, after having achieved enlightenment, would have expressed relief and contentment for living in a century in which we finally got the basis of the universe straight. Saint Paul, in his letters, echoes the same claim that the universe is finally understood. René Descartes, with the method laid out in the Discourse on the Method and Principles of Philosophy, would have made the same claim.  And so it goes, almost everyone down through history believes that he/she comprehends how the universe works. I wonder if the cow in the barn has the same illusion. Unfortunately, each has a different understanding of what it means to comprehend how the universe works, so it is not even possible to compare the relative validity of the different claims. The unconscious mind fits all it knows into a coherent framework that gives the illusion of comprehension in terms of what it considers important. In doing so, it assumes that what you see is all there is.  Kahneman refers to this as WYSIATI (What You See Is All There Is).

To a large extent the understandability of the universe is mirage based on WYSIATI—our ignorance of our ignorance. We understand as much as we are aware of and capable of understanding; blissfully ignoring the rest. We do not know how quantum gravity works, if there is intelligent life elsewhere in the universe[2], or for that matter what the weather will be like next week. While our scientific models correctly describe much about the universe, they are, in the end, only models and leave much beyond their scope, including the ultimate nature of reality.

To receive a notice of future posts follow me on Twitter: @musquod.

[1] Let’s hope time is kinder to Kahneman than it was to Freud.

[2] Given our response to global warming, one can debate if there is intelligent life on earth.

René Descartes (1596 – 1650) was an outstanding physicist, mathematician and philosopher. In physics, he laid the ground work for Isaac Newton’s (1642 – 1727) laws of motion by pioneering work on the concept of inertia. In mathematics, he developed the foundations of analytic geometry, as illustrated by the term Cartesian[1] coordinates. However, it is in his role as a philosopher that he is best remembered. Rather ironic, as his breakthrough method was a failure.

Descartes’s goal in philosophy was to develop a sound basis for all knowledge based on ideas that were so obvious they could not be doubted. His touch stone was that anything he perceived clearly and distinctly as being true was true. The archetypical example of this was the famous I think therefore I am.  Unfortunately, little else is as obvious as that famous quote and even it can be––and has been––doubted.

Euclidean geometry provides the illusionary ideal to which Descartes and other philosophers have strived. You start with a few self-evident truths and derive a superstructure built on them.  Unfortunately even Euclidean geometry fails that test. The infamous parallel postulate has been questioned since ancient times as being a bit suspicious and even other Euclidean postulates have been questioned; extending a straight line depends on the space being continuous, unbounded and infinite.

So how are we to take Euclid’s postulates and axioms?  Perhaps we should follow the idea of Sir Karl Popper (1902 – 1994) and consider them to be bold hypotheses. This casts a different light on Euclid and his work; perhaps he was the first outstanding scientist.  If we take his basic assumptions as empirical[2] rather than sure and certain knowledge, all we lose is the illusion of certainty. Euclidean geometry then becomes an empirically testable model for the geometry of space time. The theorems, derived from the basic assumption, are prediction that can be checked against observations satisfying Popper’s demarcation criteria for science. Do the angles in a triangle add up to two right angles or not? If not, then one of the assumptions is false, probably the parallel line postulate.

Back to Descartes, he criticized Galileo Galilei (1564 – 1642) for having built without having considered the first causes of nature, he has merely sought reasons for particular effects; and thus he has built without a foundation. In the end, that lack of a foundation turned out to be less of a hindrance than Descartes’ faulty one.  To a large extent, sciences’ lack of a foundation, such as Descartes wished to provide, has not proved a significant obstacle to its advance.

Like Euclid, Sir Isaac Newton had his basic assumptions—the three laws of motion and the law of universal gravity—but he did not believe that they were self-evident; he believed that he had inferred them by the process of scientific induction. Unfortunately, scientific induction was as flawed as a foundation as the self-evident nature of the Euclidean postulates. Connecting the dots between a falling apple and the motion of the moon was an act of creative genius, a bold hypothesis, and not some algorithmic derivation from observation.

It is worth noting that, at the time, Newton’s explanation had a strong competitor in Descartes theory that planetary motion was due to vortices, large circulating bands of particles that keep the planets in place.  Descartes’s theory had the advantage that it lacked the occult action at a distance that is fundamental to Newton’s law of universal gravitation.  In spite of that, today, Descartes vortices are as unknown as is his claim that the pineal gland is the seat of the soul; so much for what he perceived clearly and distinctly as being true.

Galileo’s approach of solving problems one at time and not trying to solve all problems at once has paid big dividends. It has allowed science to advance one step at a time while Descartes’s approach has faded away as failed attempt followed failed attempt. We still do not have a grand theory of everything built on an unshakable foundation and probably never will. Rather we have models of widespread utility. Even if they are built on a shaky foundation, surely that is enough.

Peter Higgs (b. 1929) follows in the tradition of Galileo. He has not, despite his Noble prize, succeeded, where Descartes failed, in producing a foundation for all knowledge; but through creativity, he has proposed a bold hypothesis whose implications have been empirically confirmed.  Descartes would probably claim that he has merely sought reasons for a particular effect: mass. The answer to the ultimate question about life, the universe and everything still remains unanswered, much to Descartes’ chagrin but as scientists we are satisfied to solve one problem at a time then move on to the next one.

To receive a notice of future posts follow me on Twitter: @musquod.

Since model building is the essence of science, this quote has a bit of a bite to it. It is from George E. P. Box (1919 – 2013), who was not only an eminent statistician but also an eminently quotable one.  Another quote from him: One important idea is that science is a means whereby learning is achieved, not by mere theoretical speculation on the one hand, nor by the undirected accumulation of practical facts on the other, but rather by a motivated iteration between theory and practice.  Thus he saw science as an iteration between observation and theory. And what is theory but the building of erroneous, or at least approximate, models?

To amplify that last comment: The main point of my philosophical musings is that science is the building of models for how the universe works; models constrained by observation and tested by their ability to make predictions for new observations, but models nonetheless. In this context, the above quote has significant implications for science. Models, even those of science, are by their very nature simplifications and as such are not one hundred per cent accurate. Consider the case of a map. Creating a 1:1 map is not only impractical[2] but even if you had one it would be one hundred per cent useless; just try folding a 1:1 scale map of Vancouver. A model with all the complexity of the original does not help us understand the original.  Indeed the whole purpose of a model is to eliminate details that are not essential to the problem at hand.

By their very nature, numerical models are always approximate and this is probably what Box had in mind with his statement. One neglects small effects like the gravitational influence of a mosquito. Even as one begins computing, one makes numerical approximations, replacing integrals with sums or vise versa, derivatives with finite differences, etc. However, one wants to control errors and keep them to a minimum. Statistical analysis techniques, such as Box developed, help estimate and control errors.

To a large extent it is self-evident that models are approximate; so what? Again to quote George Box: Since all models are wrong the scientist cannot obtain a “correct” one by excessive elaboration. On the contrary following William of Occam he should seek an economical description of natural phenomena. Just as the ability to devise simple but evocative models is the signature of the great scientist so overelaboration and overparameterization is often the mark of mediocrity. What would he have thought of a model with twenty plus parameters, like the standard model of particle physics? His point is a valid one. All measurements have experimental errors. If your fit is perfect you are almost certainly fitting noise. Hence, adding more parameters to get a perfect fit is a fool’s errand. But even without experimental error, a large number of parameters frequently means something important has been missed. Has something been missed in the standard model of particle physics with its many parameters or is the universe really that complicated?

There is an even more basic reason all models are wrong. This goes back at least as far as Immanuel Kant (1724 – 1804). He made the distinction between observation of an object and the object in itself. One never has direct experience of things, the so-called noumenal world; what one experiences is the phenomenal world as conveyed to us by our senses. What we see is not even what has been recorded by the eye.  The mind massages the raw observation into something it can understand; a useful but not necessarily accurate model of the world. Science then continues this process in a systematic manner to construct models to describe observations but not necessarily the underlying reality.

Despite being by definition at least partially wrong, models are frequently useful. The scale model map is useful to tourists trying to find their way around Vancouver or to a general plotting strategy for his next battle. But, if the maps are too far wrong the tourist will get lost and fall into False Creek and the general will go down in history as a failure. Similarly, the models for weather predictions are useful although they are certainly not a hundred per cent accurate. However, they do indicate when it safe to plan a picnic or cut the hay; provided they are right more than by chance and the standard model of particle physics, despite having many parameters and not including gravity, is a useful description of a wide range of observations. But to return to the main point, all models, even useful ones, are wrong because they are approximations and not even approximations to reality but to our observations of that reality. Where does that leave us? Well, let us save the last word for George Box: Remember that all models are wrong; the practical question is how wrong do they have to be to not be useful.