Zen and the Art of Science: A Tribute to Robert Pirsig

Author Robert Pirsig, widely acclaimed for his bestselling books, Zen and the Art of Motorcycle Maintenance (1974) and Lila (1991), passed away in his home on April 24, 2017. A well-rounded intellectual equally at home in the sciences and the humanities, Pirsig made the case that scientific inquiry, art, and religious experience were all particular forms of knowledge arising out of a broader form of knowledge about the Good or what Pirsig called “Quality.” Yet, although Pirsig’s books were bestsellers, contemporary debates about science and religion are oddly neglectful of Pirsig’s work. So what did Pirsig claim about the common roots of human knowledge, and how do his arguments provide a basis for reconciling science and religion?

Pirsig gradually developed his philosophy as response to a crisis in the foundations of scientific knowledge, a crisis he first encountered while he was pursuing studies in biochemistry. The popular consensus at the time was that scientific methods promised objectivity and certainty in human knowledge. One developed hypotheses, conducted observations and experiments, and came to a conclusion based on objective data. That was how scientific knowledge accumulated.

However, Pirsig noted that, contrary to his own expectations, the number of hypotheses could easily grow faster than experiments could test them. One could not just come up with hypotheses – one had to make good hypotheses, ones that could eliminate the need for endless and unnecessary observations and testing. Good hypotheses required mental inspiration and intuition, components that were mysterious and unpredictable.  The greatest scientists were precisely like the greatest artists, capable of making immense creative leaps before the process of testing even began.  Without those creative leaps, science would remain on a never-ending treadmill of hypothesis development – this was the “infinity of hypotheses” problem.  And yet, the notion that science depended on intuition and artistic leaps ran counter to the established view that the scientific method required nothing more than reason and the observation and recording of an objective reality.

Consider Einstein. One of history’s greatest scientists, Einstein hardly ever conducted actual experiments. Rather, he frequently engaged in “thought experiments,” imagining what it would be like to chase a beam of light, what it would feel like to be in a falling elevator, and what a clock would look like if the streetcar he was riding raced away from the clock at the speed of light.

One of the most fruitful sources of hypotheses in science is mathematics, a discipline which consists of the creation of symbolic models of quantitative relationships. And yet, the nature of mathematical discovery is so mysterious that mathematicians themselves have compared their insights to mysticism. The great French mathematician Henri Poincare believed that the human mind worked subliminally on problems, and his work habit was to spend no more than two hours at a time working on mathematics. Poincare believed that his subconscious would continue working on problems while he conducted other activities, and indeed, many of his great discoveries occurred precisely when he was away from his desk. John von Neumann, one of the best mathematicians of the twentieth century, also believed in the subliminal mind. He would sometimes go to sleep with a mathematical problem on his mind and wake up in the middle of the night with a solution. The Indian mathematical genius Srinivasa Ramanujan was a Hindu mystic who believed that solutions were revealed to him in dreams by the goddess Namagiri.

Intuition and inspiration were human solutions to the infinity-of-hypotheses problem. But Pirsig noted there was a related problem that had to be solved — the infinity of facts.  Science depended on observation, but the issue of which facts to observe was neither obvious nor purely objective.  Scientists had to make value judgments as to which facts were worth close observation and which facts could be safely overlooked, at least for the moment.  This process often depended heavily on an imprecise sense or feeling, and sometimes mere accident brought certain facts to scientists’ attention. What values guided the search for facts? Pirsig cited Poincare’s work The Foundations of Science. According to Poincare, general facts were more important than particular facts, because one could explain more by focusing on the general than the specific. Desire for simplicity was next – by beginning with simple facts, one could begin the process of accumulating knowledge about nature without getting bogged down in complexity at the outset. Finally, interesting facts that provided new findings were more important than facts that were unimportant or trivial. The point was not to gather as many facts as possible but to condense as much experience as possible into a small volume of interesting findings.

Research on the human brain supports the idea that the ability to value is essential to the discernment of facts.  Professor of Neuroscience Antonio Damasio, in his book Descartes’ Error: Emotion, Reason, and the Human Brain, describes several cases of human beings who lost the part of their brain responsible for emotions, either because of an accident or a brain tumor.  These persons, some of whom were previously known as shrewd and smart businessmen, experienced a serious decline in their competency after damage took place to the emotional center of their brains.  They lost their capacity to make good decisions, to get along with other people, to manage their time, or to plan for the future.  In every other respect, these persons retained their cognitive abilities — their IQs remained above normal and their personality tests resulted in normal scores.  The only thing missing was their capacity to have emotions.  Yet this made a huge difference.  Damasio writes of one subject, “Elliot”:

Consider the beginning of his day: He needed prompting to get started in the morning and prepare to go to work.  Once at work he was unable to manage his time properly; he could not be trusted with a schedule.  When the job called for interrupting an activity and turning to another, he might persist nonetheless, seemingly losing sight of his main goal.  Or he might interrupt the activity he had engaged, to turn to something he found more captivating at that particular moment.  Imagine a task involving reading and classifying documents of a given client.  Elliot would read and fully understand the significance of the material, and he certainly knew how to sort out the documents according to the similarity or disparity of their content.  The problem was that he was likely, all of a sudden, to turn from the sorting task he had initiated to reading one of those papers, carefully and intelligently, and to spend an entire day doing so.  Or he might spend a whole afternoon deliberating on which principle of categorization should be applied: Should it be date, size of document, pertinence to the case, or another?   The flow of work was stopped. (p. 36)

Why did the loss of emotion, which might be expected to improve decision-making by making these persons coldly objective, result in poor decision-making instead?  According to Damasio, without emotions, these persons were unable to value, and without value, decision-making in the face of infinite facts became hopelessly capricious or paralyzed, even with normal or above-normal IQs.  Damasio noted, “the cold-bloodedness of Elliot’s reasoning prevented him from assigning different values to different options, and made his decision-making landscape hopelessly flat.” (p. 51) Damasio discusses several other similar case studies.

So how would it affect scientific progress if all scientists were like the subjects Damasio studied, free of emotion, and therefore, hypothetically capable of perfect objectivity?  Well it seems likely that science would advance very slowly, at best, or perhaps not at all.  After all, the same tools for effective decision-making in everyday life are needed for the scientific enterprise as well. A value-free scientist would not only be unable to sustain the social interaction that science requires, he or she would be unable to develop a research plan, manage his or her time, or stick to a research plan.


Where Pirsig’s philosophy becomes particularly controversial and difficult to understand is in his approach to the truth. The dominant view of truth today is known as the “correspondence” theory of truth – that is, any human statement that is true must correspond precisely to something objectively real. In this view, the laws of physics and chemistry are real because they correspond to actual events that can be observed and demonstrated. Pirsig argues on the contrary that in order to understand reality, human beings must invent symbolic and conceptual models, that there is a large creative component to these models (it is not just a matter of pure correspondence to reality), and that multiple such models can explain the same reality even if they are based on wholly different principles. Math, logic, and even the laws of physics are not “out there” waiting to be discovered – they exist in the mind, which doesn’t mean that these things are bad or wrong or unreal.

There are several reasons why our symbolic and conceptual models don’t correspond literally to reality, according to Pirsig. First, there is always going to be a gap between reality and the concepts we use to describe reality, because reality is continuous and flowing, while concepts are discrete and static. The creation of concepts necessarily calls for cutting reality into pieces, but there is no one right way to divide reality, and something is always lost when this is done. In fact, Pirsig noted, our very notions of subjectivity and objectivity, the former allegedly representing personal whims and the latter representing truth, rested upon an artificial division of reality into subjects and objects; in fact, there were other ways of dividing reality that could be just as legitimate or useful. In addition, concepts are necessarily static – they can’t be always changing or we would not be able to make sense of them. Reality, however, is always changing. Finally, describing reality is not always a matter of using direct and literal language but may require analogy and imaginative figures of speech.

Because of these difficulties in expressing reality directly, a variety of symbolic and conceptual models, based on widely varying principles, are not only possible but necessary – necessary for science as well as other forms of knowledge. Pirsig points to the example of the crisis that occurred in mathematics in the nineteenth century. For many centuries, it was widely believed that geometry, as developed by the ancient Greek mathematician Euclid, was the most exact of all of the sciences.  Based on a small number of axioms from which one could deduce multiple propositions, Euclidean geometry represented a nearly perfect system of logic.  However, while most of Euclid’s axioms were seemingly indisputable, mathematicians had long experienced great difficulty in satisfactorily demonstrating the truth of one of the chief axioms on which Euclidean geometry was based. This slight uncertainty led to an even greater crisis of uncertainty when mathematicians discovered that they could reverse or negate this axiom and create alternative systems of geometry that were every bit as logical and valid as Euclidean geometry.  The science of geometry was gradually replaced by the study of multiple geometries. Pirsig cited Poincare, who pointed out that the principles of geometry were not eternal truths but definitions and that the test of a system of geometry was not whether it was true but how useful it was.

So how do we judge the usefulness or goodness of our symbolic and conceptual models? Traditionally, we have been told that pure objectivity is the only solution to the chaos of relativism, in which nothing is absolutely true. But Pirsig pointed out that this hasn’t really been how science has worked. Rather, models are constructed according to the often competing values of simplicity and generalizability, as well as accuracy. Theories aren’t just about matching concepts to facts; scientists are guided by a sense of the Good (Quality) to encapsulate as much of the most important knowledge as possible into a small package. But because there is no one right way to do this, rather than converging to one true symbolic and conceptual model, science has instead developed a multiplicity of models. This has not been a problem for science, because if a particular model is useful for addressing a particular problem, that is considered good enough.

The crisis in the foundations of mathematics created by the discovery of non-Euclidean geometries and other factors (such as the paradoxes inherent in set theory) has never really been resolved. Mathematics is no longer the source of absolute and certain truth, and in fact, it never really was. That doesn’t mean that mathematics isn’t useful – it certainly is enormously useful and helps us make true statements about the world. It’s just that there’s no single perfect and true system of mathematics. (On the crisis in the foundations of mathematics, see the papers here and here.) Mathematical axioms, once believed to be certain truths and the foundation of all proofs, are now considered definitions, assumptions, or hypotheses. And a substantial number of mathematicians now declare outright that mathematical objects are imaginary, that particular mathematical formulas may be used to model real events and relationships, but that mathematics itself has no existence outside the human mind. (See The Mathematical Experience by Philip J. Davis and Reuben Hersh.)

Even some basic rules of logic accepted for thousands of years have come under challenge in the past hundred years, not because they are absolutely wrong, but because they are inadequate in many cases, and a different set of rules is needed. The Law of the Excluded Middle states that any proposition must be either true or false (“P” or “not P” in symbolic logic). But ever since mathematicians discovered propositions which are possibly true but not provable, a third category of “possible/unknown” has been added. Other systems of logic have been invented that use the idea of multiple degrees of truth, or even an infinite continuum of truth, from absolutely false to absolutely true.

The notion that we need multiple symbolic and conceptual models to understand reality remains controversial to many. It smacks of relativism, they argue, in which every person’s opinion is as valid as another person’s. But historically, the use of multiple perspectives hasn’t resulted in the abandonment of intellectual standards among mathematicians and scientists. One still needs many years of education and an advanced degree to obtain a job as a mathematician or scientist, and there is a clear hierarchy among practitioners, with the very best mathematicians and scientists working at the most prestigious universities and winning the highest awards. That is because there are still standards for what is good mathematics and science, and scholars are rewarded for solving problems and advancing knowledge. The fact that no one has agreed on what is the One True system of mathematics or logic isn’t relevant. In fact, physicist Stephen Hawking has argued:

[O]ur brains interpret the input from our sensory organs by making a model of the world. When such a model is successful at explaining events, we tend to attribute to it, and to the elements and concepts that constitute it, the quality of reality or absolute truth. But there may be different ways in which one could model the same physical situation, with each employing different fundamental elements and concepts. If two such physical theories or models accurately predict the same events, one cannot be said to be more real than the other; rather we are free to use whichever model is more convenient (The Grand Design, p. 7).

Among the most controversial and mind-bending claims Pirsig makes is that the very laws of nature themselves exist only in the human mind. “Laws of nature are human inventions, like ghosts,” he writes. Pirsig even remarks that it makes no sense to think of the law of gravity existing before the universe, that it only came into existence when Isaac Newton thought of it. It’s an outrageous claim, but if one looks closely at what the laws of nature actually are, it’s not so crazy an argument as it first appears.

For all of the advances that science has made over the centuries, there remains a sharp division of views among philosophers and scientists on one very important issue: are the laws of nature actual causal powers responsible for the origins and continuance of the universe or are the laws of nature summary descriptions of causal patterns in nature? The distinction is an important one. In the former view, the laws of physics are pre-existing or eternal and possess god-like powers to create and shape the universe; in the latter view, the laws have no independent existence – we are simply finding causal patterns and regularities in nature that allow us to predict and we call these patterns “laws.”

One powerful argument in favor of the latter view is that most of the so-called “laws of nature,” contrary to the popular view, actually have exceptions – and sometimes the exceptions are large. That is because the laws are simplified models of real phenomena. The laws were cobbled together by scientists in order to strike a careful balance between the values of scope, predictive accuracy, and simplicity. Michael Scriven, a mathematician and philosopher at Claremont Graduate University, has noted that as a result of this balance of values, physical laws are actually approximations that apply only within a certain range. This point has also been made more recently by Ronald Giere, a professor of philosophy at the University of Minnesota, in Science Without Laws and Nancy Cartwright of the University of California at San Diego in How the Laws of Physics Lie.

Newton’s law of universal gravitation, for example, is not really universal. It becomes increasingly inaccurate under conditions of high gravity and very high velocities, and at the atomic level, gravity is completely swamped by other forces. Whether one uses Newton’s law depends on the specific conditions and the level of accuracy one requires. Newton’s laws of motion also have exceptions, depending on the force, distance, and speed. Kepler’s laws of planetary motion are an approximation based on the simplifying assumption of a planetary system consisting of one planet. The ideal gas law is an approximation which becomes inaccurate under conditions of low temperature and/or high pressure. The law of multiple proportions works for simple molecular compounds, but often fails for complex molecular compounds. Biologists have discovered so many exceptions to Mendel’s laws of genetics that some believe that Mendel’s laws should not even be considered laws.

So if we think of laws of nature as being pre-existing, eternal commandments, with god-like powers to shape the universe, how do we account for these exceptions to the laws? The standard response by scientists is that their laws are simplified depictions of the real laws. But if that is the case, why not state the “real” laws? Because by the time we wrote down the real laws, accounting for every possible exception, we would have an extremely lengthy and detailed description of causation that would not recognizably be a law. The whole point of the laws of nature was to develop tools by which one could predict a large number of phenomena (scope), maintain a good-enough correspondence to reality (accuracy), and make it possible to calculate predictions without spending an inordinate amount of time and effort (simplicity). That is why although Einstein’s conception of gravity and his “field equations” have supplanted Newton’s law of gravitation, physicists still use Newton’s “law” in most cases because it is simpler and easier to use; they only resort to Einstein’s complex equations when they have to! The laws of nature are human tools for understanding, not mathematical gods that shape the universe. The actual practice of science confirms Pirsig’s point that the symbolic and conceptual models that we create to understand reality have to be judged by how good they are – simple correspondence to reality is insufficient and in many cases is not even possible anyway.



Ultimately, Pirsig concluded, the scientific enterprise is not that different from the pursuit of other forms of knowledge – it is based on a search for the Good. Occasionally, you see this acknowledged explicitly, when mathematicians discuss the beauty of certain mathematical proofs or results, as defined by their originality, simplicity, ability to solve many problems at once, or their surprising nature. Scientists also sometimes write about the importance of elegance in their theories, defined as the ability to explain as much as possible, as clearly as possible, and as simply as possible. Depending on the field of study, the standards of judgment may be different, the tools may be different, and the scope of inquiry is different. But all forms of human knowledge — art, rhetoric, science, reason, and religion — originate in, and are dependent upon, a response to the Good or Quality. The difference between science and religion is that scientific models are more narrowly restricted to understanding how to predict and manipulate natural phenomena, whereas religious models address larger questions of meaning and value.

Pirsig did not ignore or suppress the failures of religious knowledge with regard to factual claims about nature and history. The traditional myths of creation and the stories of various prophets were contrary to what we know now about physics, biology, paleontology, and history. In addition, Pirsig was by no means a conventional theist — he apparently did not believe that God was a personal being who possessed the attributes of omniscience and omnipotence, controlling or potentially controlling everything in the universe.

However, Pirsig did believe that God was synonymous with the Good, or “Quality,” and was the source of all things.  In fact, Pirsig wrote that his concept of Quality was similar to the “Tao” (the “Way” or the “Path”) in the Chinese religion of Taoism. As such, Quality was the source of being and the center of existence. It was also an active, dynamic power, capable of bringing about higher and higher levels of being. The evolution of the universe, from simple physical forms, to complex chemical compounds, to biological organisms, to societies was Dynamic Quality in action. The most recent stage of evolution – Intellectual Quality – refers to the symbolic models that human beings create to understand the universe. They exist in the mind, but are a part of reality all the same – they represent a continuation of the growth of Quality.

What many religions were missing, in Pirsig’s view, was not objectivity, but dynamism: an ability to correct old errors and achieve new insights. The advantage of science was its willingness and ability to change. According to Pirsig,

If scientists had simply said Copernicus was right and Ptolemy was wrong without any willingness to further investigate the subject, then science would have simply become another minor religious creed. But scientific truth has always contained an overwhelming difference from theological truth: it is provisional. Science always contains an eraser, a mechanism whereby new Dynamic insight could wipe out old static patterns without destroying science itself. Thus science, unlike orthodox theology, has been capable of continuous, evolutionary growth. (Lila, p. 222)

The notion that religion and orthodoxy go together is widespread among believers and secularists. But there is no necessary connection between the two. All religions originate in social processes of story-telling, dialogue, and selective borrowing from other cultures. In fact, many religions begin as dangerous heresies before they become firmly established — orthodoxies come later. The problem with most contemporary understandings of religion is that one’s adherence to religion is often measured by one’s commitment to orthodoxy and membership in religious institutions rather than an honest quest for what is really good.  A person who insists on the literal truth of the Bible and goes to church more than once a week is perceived as being highly religious, whereas a person not connected with a church but who nevertheless seeks religious knowledge wherever he or she can find it is considered less committed or even secular.  This prejudice has led many young people to identify as “spiritual, not religious,” but religious knowledge is not inherently about unwavering loyalty to an institution or a text. Pirsig believed that mysticism was a necessary component of religious knowledge and a means of disrupting orthodoxies and recovering the dynamic aspect of religious insight.

There is no denying that the most prominent disputes between science and religion in the last several centuries regarding the physical workings of the universe have resulted in a clear triumph for scientific knowledge over religious knowledge.  But the solution to false religious beliefs is not to discard religious knowledge — religious knowledge still offers profound insights beyond the scope of science. That is why it is necessary to recover the dynamic nature of religious knowledge through mysticism, correction of old beliefs, and reform. As Pirsig argued, “Good is a noun.” Not because Good is a thing or an object, but because Good  is the center and foundation of all reality and all forms of knowledge, whether we are consciously aware of it or not.

What Are the Laws of Nature? – Part Two

In a previous post, I discussed the mysterious status of the “laws of nature,” pointing out that these laws seem to be eternal, omnipresent, and possessing enormous power to shape the universe, although they have no mass and no energy.

There is, however, an alternative view of the laws of nature proposed by thinkers such as Ronald Giere and Nancy Cartwright, among others. In this view, it is a fallacy to suppose that the laws of nature exist as objectively real entities — rather, what we call the laws of nature are simplified models that the human mind creates to explain and predict the operations of the universe. The laws were created by human beings to organize information about the cosmos. As such, the laws are not fully accurate descriptions of how the universe actually works, but generalizations; and like nearly all generalizations, there are numerous exceptions when the laws are applied to particular circumstances. We retain the generalizations because they excel at organizing and summarizing vast amounts of information, but we should never make the mistake of assuming that the generalizations are real entities. (See Science Without Laws and How the Laws of Physics Lie.)

Consider one of the most famous laws of nature, Isaac Newton’s law of universal gravitation. According to this law, the gravitational relationship between any two bodies in the universe is determined by the size (mass) of the two bodies and their distance from each other. More specifically, any two bodies in the universe attract each other with a force that is (1) directly proportional to the product of their masses and (2) inversely proportional to the square of the distance between them.  The equation is quite simple:

F = G \frac{m_1 m_2}{r^2}\

where F is the force between two masses, G is a gravitational constant, m1 and m2 are the masses of the two bodies and r is the distance between the center of the two bodies.

Newton’s law was quite valuable in helping predict the motions of the planets in our solar system, but in some cases the formula did not quite match to astronomical observations. The orbit of the planet Mercury in particular never fit Newton’s law, no matter how much astronomers tried to fiddle with the law to get the right results. It was only when Einstein introduced his theory of relativity that astronomers could correctly predict the motions of all the planets, including Mercury. Why did Einstein’s theory work better for Mercury? Because as the planet closest to the sun, Mercury is most affected by the massive gravitation of the sun, and Newton’s law becomes less accurate under the conditions of massive gravitation.

Einstein’s equations for gravity are known as the “field equations,” and although they are better at predicting the motions of the planets, they are extremely complex — too complex really for many situations. In fact, physicist Stephen Hawking has noted that scientists still often use Newton’s law of gravity because it is much simpler and a good enough approximation in most cases.

So what does this imply about the reality of Newton’s law of universal gravitation? Does Newton’s law float around in space or in some transcendent realm directing the motions of the planets, until the gravitation becomes too large, and then it hands off its duties to the Einstein field equations? No, of course not. Newton’s law is an approximation that works for many, but not all cases. Physicists use it because it is simple and “good enough” for most purposes. When the approximations become less and less accurate, a physicist may switch to the Einstein field equations, but this is a human value judgment, not the voice of nature making a decision to switch equations.

One other fact is worth noting: in Newton’s theory, gravity is a force between two bodies. In Einstein’s theory, gravity is not a real force — what we call a gravitational force is simply how we perceive the distortion of the space-time fabric caused by massive objects. Physicists today refer to gravity as a “fictitious force.” So why do professors of physics continue to use Newton’s law and teach this “fictitious force” law to their students? Because it is simpler to use and still a good enough approximation for most cases. Newton’s law can’t possibly be objectively real — if it is, Einstein is wrong.

The school of thought known as “scientific realism” would dispute these claims, arguing that even if the laws of nature as we know them are approximations, there are still real, objective laws underneath these approximations, and as science progresses, we are getting closer and closer to knowing what these laws really are. In addition, they argue that it would be absurd to suppose that we can possibly make progress in technology unless we are getting better and better in knowing what the true laws are really like.

The response of Ronald Giere and Nancy Cartwright to the realists is as follows: it’s a mistake to assume that if our laws are approximations and our approximations are getting better and better that therefore there must be real laws underneath. What if nature is inherently so complex in its causal variables and sequences that there is no objectively real law underneath it all? Nancy Cartwright notes that engineers who must build and maintain technological devices never apply the “laws of nature” directly to their work without a great deal of tinkering and modifications to get their mental models to match the specific details of their device. The final blueprint that engineers may create is a highly specific and highly complex model that is a precise match for the device, but of very limited generalizability to the universe as a whole. In other words, there is an inherent and unavoidable tradeoff between explanatory power and accuracy. The laws of nature are valued by us because they have very high explanatory power, but specific circumstances are always going to involve a mix of causal forces that refute the predictions of the general law. In order to understand how two bodies behave, you not only need to know gravity, you need to know the electric charge of the two bodies, the nuclear force, any chemical forces, the temperature, the speed of the objects, and additional factors, some of which can never be calculated precisely. According to Cartwright,

. . . theorists tend to think that nature is well-regulated; in the extreme, that there is a law to cover every case. I do not. I imagine that natural objects are much like people in societies. Their behavior is constrained by some specific laws and by a handful of general principles, but it is not determined in detail, even statistically. What happens on most occasions is dictated by no law at all. . . . God may have written just a few laws and grown tired. We do not know whether we are living in a tidy universe or an untidy one. (How the Laws of Physics Lie, p. 49)

Cartwright makes it clear that she believes in causal powers in nature — it’s just that causal powers are not the same as laws, which are simply general principles for organizing information.

Some philosophers and scientists would go even further. They argue that science is able to develop and improve models for predicting phenomena, but the underlying nature of reality cannot be grasped directly, even if our models are quite excellent at predicting. This is because there are always going to be aspects of nature that are non-observable and there are often multiple theories that can explain the same phenomenon. This school of thought is known as instrumentalism.

Stephen Hawking appears to be sympathetic to such a view. In a discussion of his use of “imaginary time” to model how the universe developed, Hawking stated “a scientific theory is just a mathematical model we make to describe our observations: it exists only in our minds. So it is meaningless to ask: which is real, “real” or “imaginary” time? It is simply a matter of which is the more useful description.” (A Brief History of Time, p. 144) In a later essay, Hawking made the case for what he calls “model-dependent realism.” He argues:

it is pointless to ask whether a model is real, only whether it agrees with observation. If two models agree with observation, neither one can be considered more real than the other. A person can use whichever model is more convenient in the situation under consideration. . . . Each theory may have its own version of reality, but according to model-dependent realism, that diversity is acceptable, and none of the versions can be said to be more real than any other.

Hawking concludes that given these facts, it may well be impossible to develop a unified theory of everything, that we may have to settle for a diversity of models. (It’s not clear to me how Hawking’s “model-dependent realism” differs from instrumentalism, since they seem to share many aspects.)

Intuitively, we are apt to conclude that our progress in technology is proof enough that we are understanding reality better and better, getting closer and closer to the Truth. But it’s actually quite possible for science to develop better and better predictive models while still retaining very serious doubts and disputes about many fundamental aspects of reality. Among physicists and cosmologists today, there is still disagreement on the following issues: are there really such things as subatomic particles, or are these entities actually fields, or something else entirely?; is the flow of time an illusion, or is time the chief fundamental reality?; are there an infinite number of universes in a wider multiverse, with infinite versions of you, or is this multiverse theory a mistaken interpretation of uncertainty at the quantum level?; are the constants of the universe really constant, or do they sometimes change?; are mathematical objects themselves the ultimate reality, or do they exist only in the mind? A number of philosophers of science have concluded that science does indeed progress by creating more and better models for predicting, but they make an analogy to evolution: life forms may be advancing and improving, but that doesn’t mean they are getting closer and closer to some final goal.

Referring back to my previous post, I discussed the view that the “laws of nature” appear to exist everywhere and have the awesome power to shape the universe and direct the motions of the stars and planets, despite the fact that the laws themselves have no matter and no energy. But if the laws of nature are creations of our minds, what then? I can’t prove that there are no real laws behind the mental models that we create. It seems likely that there must be some such laws, but perhaps they are so complex that the best we can do is create simplified models of them. Or perhaps we must acknowledge that the precise nature of the cosmological order is mysterious, and any attempt to understand and describe this order must use a variety of concepts, analogies, and stories created by our minds. Some of these concepts, analogies, and stories are clearly better than others, but we will never find one mental model that is a perfect fit for all aspects of reality.