The Science of Demand (4) - Unofficial Translation of Steven Cheung's 经济解释 - 科学说需求

Section 4.  The Significance of Refutability by Facts

If the reader asks: in the whole structure of scientific methodology, which part is the most significant? With no hesitation, my answer is: theoretical prediction must be “refutable by facts”. A theory that is not refutable by facts has no explanatory power. It can be said that the subject of all empirical science is to establish certain refutable statement for prediction. In other words, science does not seek to be right, nor seek to be wrong; science seeks to be “refutable by facts”. Potentially refutable by facts but has not been refuted can then be said to have been confirmed. As aforementioned, to predict the occurrence of a phenomenon and to explain a phenomenon is the same thing. Prediction refutable by a phenomenon but has not been so refuted, whereas the occurrence of a phenomenon confirms the prediction, such phenomenon can then be said to have been explained. Certainly, a phenomenon can be explained by a number of theories. I will subsequently touch upon the choice of different theories.

The key here is: the reason why a theory not refutable by facts has no explanatory power is that such a theory cannot be empirically tested. A tautology can never be falsified, and since it can never be falsified, how can it be refuted by facts? A theory refutable by facts must be capable of being falsified by imagination. A tautology, being non-falsifiable even by imagination, has no explanatory power. Besides tautologies, there are four other situations making theories not refutable by facts, hence destroying the explanatory power of these theories. These are the contents of Sections 5 and 6 that follow.

Being refutable by facts is important. If theoretical prediction is refuted by facts, there are only two choices. The first one is to abandon the theory and look for others; the second one is to insert conditions to remedy it. But as mentioned when discussing ad hoc theory, such remedy requires option to be forgone, and such option forgone should not be too much. Refutable theory that has not been refuted today may not remain so tomorrow – this is the course of scientific advancement. Not refuted as of today means that the theory still retains its uses. The most important criterion in valuing a theory lies in its applicability to explain phenomenon, not in its right or wrong.

Prediction using a sentence or a statement has to focus on testable or refutable implication. Such implication is derived from theory. Logically, the rules for implication are simple: if the occurrence of A implies the occurrence of B (A → B), then the non-occurrence of B implies the non-occurrence of A (Not B → Not A). This is the most fundamental test. For instance, if it rains (A), then there are clouds in the sky (B). It implies: if there are no clouds (Not B), then it does not rain (Not A). If there are no clouds but it rains, then the hypothesis that if it rains (A), there are clouds (B) is refuted by facts.

The way to test the implication of a theory is to provide evidence to the contrary. This is a crucial point. In testing the implication that if it rains, then there are clouds (A → B), evidence to the contrary that if there are no clouds, then it does not rain (Not B → Not A) has to be provided. To test using if it does not rain, then there are no clouds (Not A → Not B) is a common fallacy (in logic, such fallacy is termed fallacy of denying the antecedent). The occurrence of A implies the occurrence of B, but the non-occurrence of A has no implication whatsoever on B. It is a common misconception to say the non-occurrence of A implies the non-occurrence of B, with many scholars falling into this trap. For instance, in economics we assume every individual maximizes his self-interest (A), therefore under certain constraints, every individual will work hard (B). Some scholars submit that an individual does not necessarily maximize his self-interest (Not A), therefore under the same constraints, an individual will not necessarily work hard (Not B). This is a fallacy.

In 1946, an economist called R. A. Lester published an article catching the whole world’s attention. After investigating the policy of hiring drivers in Boston’s private transport companies, he declared wrong the renowned “marginal productivity theory” (the word “marginal” is not important here, but will later be expounded). According to economic postulate, every private company will maximize its profit, therefore in hiring drivers for its trucks, the marginal productivity contribution of every driver will equal his wage (this is one implication of the “marginal productivity theory”). Lester checked with all the principals of Boston’s transport companies and found that they neither knew nor cared what “marginal productivity” was, therefore this theorem was wrong: the wages of drivers did not equal their marginal productivity contribution. This fell into the aforementioned misconception that if it does not rain, then there are no clouds.

I can cite an interesting (but non-factual) example to illustrate this “A → B, therefore Not A → Not B” fallacy. Let’s say there were a group of idiots who knew nothing about real-world phenomena. Yet economists assumed that they all wisely maximized their self-interest. In fact, these people were all idiots, therefore this assumption was clearly wrong. On hearing that gasoline station was fun, these idiots all set up their own gasoline stations. Being idiots, some built their gasoline stations on remote mountains, some in dense forests, while some others on the sea. With no highways for vehicles to pass through, how could these gasoline stations survive? Nonetheless, some idiots in the same group unknowingly built their gasoline stations along highways. Not long afterward witnessed the survival of the fittest. Only those stations situating along highways survived. In fact, the idiots had not had a clue of what they were doing. The economists’ assumption that they knew how to maximize their self-interests was obviously wrong. Yet the surviving gasoline stations coincided with the postulate of maximization of self-interest, such behavior was therefore explained. Reasoning that since the idiots had no idea what they did, gasoline stations would not be built on a profitable location is fallacious.

Astronomical science in ancient China was brilliant, having high accuracy in predicting the timings of eclipses and lunar eclipses. I have not gone into any depth exploring what this science is all about, but kids in China have heard about the legend of “heavenly dog swallowing the sun”. Using the postulate of “heavenly dog” to explain eclipse or lunar eclipse is certainly nonsense, but if prediction is accurate and not refuted by facts, it might as well be acceptable. Today’s theories on eclipse and lunar eclipse having replaced those of ancient times is not because today’s are right while yesterday’s were wrong, but due to today’s having more generalization to explain other astronomical phenomena. Perhaps when tomorrow comes, today’s theories could be proven wrong. Tautology is absolute yet without explanatory power. A theory with explanatory power could be wrong, but more importantly, it is refutable by facts. No matter right or wrong, a theory with explanatory power is a useful one. Reasoning that there is no heavenly dog swallowing the sun, therefore it cannot be used to predict the occurrence of eclipse is a fallacy. It is paramount to make the issue crystal-clear.

The Science of Demand (3) - Unofficial Translation of Steven Cheung's 经济解释 - 科学说需求

Section 3.  Ad hoc Theory and Tautology

We know that the same piece of good will weigh less if it is up on a high mountain. The law of gravity explains this phenomenon. Yet before Newton, what did people think? We know that temperature drops on a high mountain, therefore we say, low temperatures, for some reason, reduce the weight of goods. This is a theory. To verify the theory, we put the same piece of good down to sea level in a refrigerated room, measure its weight and find that it does not weigh any less. This theory between temperature and weight is thus refuted.

Subsequently I will explain that for every theory with explanatory power, it must be refutable by facts but has not been so refuted. Using the reduction of temperature to explain the reduction of weight has been refuted by facts, so should we take this as wrong? This is an important philosophical question.

If regardless of other circumstances, any theory refuted by facts is taken as wrong, then all theories are wrong. This is not acceptable. Theories refuted by facts can be remedied. Using the weight of a good on a mountain as an example, the rationale of falling temperature has been overthrown, we can say instead that not only is temperature lower on a mountain, but wind is also stronger. Therefore, we carry out another experiment, putting the same piece of good in a freezing compartment with the addition of a blowing fan to measure its weight. Such measurement would reveal again that the temperature hypothesis is wrong.

Without giving up, we also note that the hillside slopes. Therefore, on top of a freezing compartment and a blowing fan, we add a tilted board and put on it the piece of good to measure the good’s weight. Again, the temperature hypothesis is not credible. Not discouraged, we point out that a high mountain is way above sea level. Therefore, we spend a great deal of money building a sky-high freezing compartment. Eventually, we duplicate the conditions of a high mountain: with temperature freezing cold; with wind from a blowing fan; with slope by a tilted board; with height sky-high, a piece of good really weighs less. The temperature hypothesis is therefore confirmed. This theory is not incorrect, yet it is an ad hoc theory. An ad hoc theory, also a theory but since it is too ad hoc, has no generalized explanatory power. It is not due to a lack of content of the theory. On the contrary, it has too much content, therefore when the content is slightly altered, the theory is overthrown.

Any scientific theory, even if refuted by facts, can be remedied by incorporating more conditions. But option has to be forgone in remedying a theory. Too much option forgone is not warranted. Option forgone will be too much if an ad hoc theory explains only a single phenomenon but cannot be extended to theorizing other phenomena; has no generalization function; and minimal explanatory power. Theories refuted by facts can be remedied, and often should be remedied, yet option forgone should not be too much. The guideline in measuring whether option forgone is too much is based on the magnitude of explanatory power. We should not abandon a theory when its explanatory power is not extensive – a non-extensive theory today may be replaced by another with more extensive explanatory power tomorrow, but before that happens, a non-extensive theory could have already been the most useful.

There is unalterable truth in the world, yet any theory is replaceable by a better theory. Scientific advancement is not due to a correct theory replacing an incorrect one, but due to a theory with more extensive explanatory power replacing a less extensive one. Advances in human thought can render what is considered superb today replaced by another with more applications tomorrow. We are yet to put a full stop to the capability of human thought. Science progresses by leaps and bounds since World War II, giving us reasons to believe that human thought may have no boundary.

If an ad hoc theory is so specific as to explain only a single phenomenon – like the aforementioned example that explains only the weight of an object on a high mountain – it stands at one extreme of scientific theory which has minimal application which cannot at all be generalized. Theories at the other extreme, however, can be outrageously generalized so that they can never be falsified under any circumstances. They cannot be wrong because they are devoid of content. This is what philosophy terms tautology. An ad hoc theory has too much content while a tautology has none. A commendable theory must lie somewhere between an ad hoc theory and a tautology.

The so-called tautology refers to certain statement that cannot be falsified under any circumstances. In a stricter sense, a tautological statement cannot be conceived to be wrong! For instance, if I say: “A four-leg animal has four legs.” How can this be possibly wrong? The second part of the sentence reiterates the meaning of the first. Even if we spend loads of effort, under no circumstances can this be conceived to be wrong. It cannot be wrong on earth, on Mars, or anywhere within the universe. This sentence possesses powerful generalization, but what is its content? None has it in fact! No matter how hard we deliberate, we know this is correct, yet we do not know its content. A tautology is empty with zero explanatory power.

In general, a tautological statement is not as trivial as “a four-leg animal has four legs” that can be identified at a glance. “Theories” which carry no substance and cannot be falsified are aplenty, yet very often are not detected even by scholarly doctors. Let me cite a few examples.

An indispensable postulate in economics is: every behavior of every individual is for maximizing self-interest. However, a person hurts himself if he smokes or jumps off a building. If we say smoking or plunging from a building is because of “maximization of self-interest”, this is a tautological statement. With all behavior counted, using this postulate of “maximization of self-interest” to “explain” smoking or plunging from a building cannot be falsified, since the postulate itself generally incorporates all behavior of an individual. If the behavior of every individual could be explained by definition and in such an empty manner, then the entire economics would barely have any content.

Let’s quote another example. An economist attempted to empirically test whether a private enterprise’s production cost was the lowest possible of that enterprise. By economic definition, in order to maximize profits, all private enterprises will do their utmost to lower their production costs. Therefore, the hypothesis of this economist was a tautological statement. It could never be falsified, but it carried no content as the definition itself does not allow any behavior of intentionally not lowering costs when presented with such an opportunity. Friedman gave some remarkable comment on this economist’s empirical work: “Stupid question will of course yield stupid answer!” What is a stupid question? A question that cannot possibly have a second answer – or a question that cannot possibly have a wrong answer – is stupid.

A tautology is not necessarily superficial. Very often it cannot be discovered at a glance, and at times not even by learned scholar. More than forty years ago, a Harvard University graduate was awarded a Ph.D. in economics, with his dissertation winning an excellence award. That dissertation was later published in a book and vigorously trumpeted. Even more well-known was the book review by Armen Alchian. Alchian brilliantly pointed out that the whole dissertation was a tautology, devoid of content and could not be falsified. That book review deeply embarrassed Harvard. Just imagine that even top-notch economics professors in the renowned Harvard University could not discover the tautology of a Ph.D. student, how can we underestimate the “profundity” of this kind of logic?

I said a tautological statement cannot be falsified, carries no content, but did not say such a statement could not possibly be an important concept. In fact, many important scientific theories originated from the viewpoints or concepts of tautologies. There is a commendable feature of tautology: it can be vastly generalized. If we can restrain or limit its scope, sometimes a falsifiable theory with content can be devised. Its explanatory power could be so strong as to win lots of plaudits.

We can quote a few examples in economics. It is a tautology devoid of content if the aforementioned “maximization of self-interest” and smoking are mixed up, like by definition, with seemingly perfect justification. But if we can insert a few constraints to enable us to infer under what circumstances a person would smoke more, smoke less, or quit smoking, then such a theory has content to be empirically tested.

A more distinct example, turning a tautology into a theory with broad applications, is the quantity theory of money in monetary theory. The starting point of this theory is obviously a tautology: money supply (M) times the circulating velocity of money (V) equals the price of goods (P) times the transacted quantity of goods (Q). Such an MV = PQ equation cannot be falsified, as the former (MV) and the latter (PQ) are merely different perspectives of the same amount. Since this equation cannot be falsified, it becomes a definition and can thus be written as MV = PQ. Clearly this definition does not explain anything, but since it provides a new perspective to look at the world, it is inspiring. When appropriately restrained, it becomes the important quantity theory of money with massive explanatory power. Extensively learned scholars like Irving Fisher and Friedman successfully indicated under what circumstances the circulating velocity of money (V) would be roughly constant, then went on to specify the relationship between money supply (M) and the price of goods (P). The quantity theory can be amazingly applied to ever-changing situations. Ultimately, its origin was a tautological concept.

Some people say that the Coase theorem, vastly popular in economics for over forty years, is a tautology. I consider the Coase theorem immensely useful since those knowledgeable can skillfully insert constraints to generate many hypotheses capable of explaining different phenomena. In the hands of people with varying abilities, the same tautology could yield distinctly different clout. Those who criticize the Coase theorem as a tautology and turn a blind eye to it have no idea of its immensity. As to what the Coase theorem is, we will analyze in detail in Volume III.

We can draw some conclusion between the two extremes of ad hoc theory and tautology. An ad hoc theory has too much content, can specifically explain a single phenomenon but its explanatory power cannot be generalized. Yet having an ad hoc theory is nonetheless better than having no theory at all. As well said by Reuben Kessel: “No argument can be won with no theoretical underpinning.” The capacity to explain a single phenomenon is better than the incapacity to explain any phenomenon, though any commendable scientific theory must be capable of generalization; otherwise theories could be as plentiful as phenomena, and the world would then be a big mess. 

The other extreme is: since a tautology is too general and cannot be falsified, its content tends to be empty and irrelevant. The explanatory power of a tautology is even weaker than that of an ad hoc theory, yet a tautology can be an inspiringly important concept in providing us a new perspective to view the world. Those who believe a tautology is devoid of content and turn a blind eye to it could have given up a treasure. Instead of abandoning a new perspective to view the world, we should try to incorporate constraints to add content to a tautology, hoping to turn “definition” into a theory capable of explaining phenomena.

Greatly commendable theories capable of explaining phenomena always lie somewhere between the two extremes of ad hoc theory and tautology. Scientific advancement often commences from one extreme or the other and evolves progressively toward the center.

The Science of Demand (2) - Unofficial Translation of Steven Cheung's 经济解释 - 科学说需求

Section 2.  Facts Not Explainable by Facts

In science, phenomenon, fact, behavior or observation refer to the same thing – though certain phenomena cannot be physically observed.

Explaining a phenomenon often requires the use of non-factual abstract theory. Why should the explanation of a fact involve abstract thinking? The answer is that the regularity of a fact is not self-explanatory. If it rains, there are clouds – this is a regular phenomenon – but the fact that it rains cannot explain the existence of clouds. Wheat grows in soil – this is regularity – but soil cannot explain wheat. The delineation of rights brings economic prosperity – this is regularity, too – but economic prosperity cannot explain the existence of the delineation of property rights; neither is there any explanatory power the other way round. The regularity of facts could only tell us how, but not why.

Suppose phenomenon A leads to phenomenon B, and then we say A explains B, or B explains A, there are two problems. First, there are numerous regular phenomena in the world. If they are really self-explanatory, then there would be voluminous theories in each and every science, and none of them would have any generalized explanatory power. If one phenomenon can explain another phenomenon, then as long as the regularity of a phenomenon is discovered, and we consider the regularity self-explanatory, what would remain in human’s line of reasoning? Second, for a phenomenon exhibiting regularity, the behavior of the regularity may change under different circumstances. For instance, a feather should fall, but may rise when wind is blowing. If wind is used to explain why a feather rises, then why doesn’t a rock ascend under windy condition? What principles should be applied for categorization? The principle we are looking for is a scientific tenet or theory. One application of science, we might say, is to systematically arrange and categorize phenomena.

Karl Brunner said: “Facts cannot be used to explain facts.” Milton Friedman said: “Regularities of facts have to be explained.” In the economics profession, the best quote comes from Alfred Marshall: “All the controversies tell us that unless filtered by rational investigation and interpretation, it is impossible for us to learn anything from facts. This also teaches us that the most reckless and hypocritical people are those theorists who openly profess to let facts do their self-explanation; or unconsciously control the selection and combination of facts behind the scenes, and then put forward the following prediction: since that follows this, this is therefore the reason.”

The Science of Demand (1) - Unofficial Translation of Steven Cheung's 经济解释 - 科学说需求

More than ten years ago, while wandering in a bookshop, I incidentally came across a book of Steven N S Cheung. Once started reading, there was no way to stop. A whole decade of collecting and studying almost every publication of Cheung followed. Some are not easily comprehensible, while most are amazing and fun. About three years ago, after learning Cheung was drastically revising his classic Chinese book “Economic Explanation”, I volunteered to help translate that into English. The offer was unfortunately turned down. Rules, however, cannot deter follies – I spent several months translating the then newly published Volume I into simple English, then e-mailed the script to Cheung for review. Two years have gone by with no response. Understandably, Cheung is still focusing on rewriting the remainder of his classic, yet follies are nothing less than follies – in the sheer hope of spreading sooner Cheung’s lifelong economics thinking into the English world, the translation will be published in this blog. Treasured knowledge should be shared in the public domain, instead of being barred by language barrier!

Economic Explanation     Volume I     The Science of Demand
Written by: Steven N S Cheung     Translated by: Daley Mok

Chapter I.  Scientific Methodology

Sitting by my desk, with a pen in hand, I am thinking about mankind’s achievements in science. Science is an interesting knowledge in systematically explaining phenomena. Definitely, mankind is the crown of creation. So developed is our brain that when compared with other creatures, the difference is immense. Expression of feeling is an art, whereas rational analysis is a science. Human feelings are often confused with reasoning. As such, scientific inference may be influenced by feelings and thus made sloppy, or it can be so fantastically made that words could not express. Indeed, science may have the beauty of art.

In search of beauty is only human nature, science therefore may have its artistic quality and beauty, too. However, science itself is no art, its main objective being to explain phenomena. On the other hand, human, being only human, cannot be so cold-hearted as to be emotionless. Consequently, declaring a scientific article as a piece of art is a compliment. The problem is, merely beautiful but incapable of explaining phenomena means that science loses its function. Since scientists are also only human, we cannot expect they are any exceptions with no emotions. However, emotions can never be abused in science. The principle is simple: scientific work can combine objective analysis and subjective judgment, yet the two have to be clearly differentiated. As long as this is done, affectionate terms can be inserted as ornaments in scientific writings to make them less monotonous but more readable.

In economics, subjective sentiment and objective analysis are more difficult to delineate. Though not impossible, it is more difficult when compared to natural sciences such as physics or chemistry. Economics is a science for explaining human behavior. The problem is, economists are only human, thus inevitably add their own values, and may even apply their own likes and dislikes to scientific conclusion. First-rate economists, however, are capable of setting aside their own value judgment when analyzing. This is the ability to simultaneously use one’s mind on two separate matters. People who are not born with this have to work extra hard.

Section 1.  Regularity of Phenomenon

My desk sits next to the window. It is late autumn. Bamboos outside are swaying under the breeze. In a densely populated city like Hong Kong, it is not easy to have a tree-lined window view. Du Fu, our great poet, wrote: “Carpeting the ground are leaves cascading down from boundless trees.” Hong Kong people could envisage such a scene despite not experiencing that before. Why? It is already late autumn, yet the greenness of the bamboos here are still lovely. Why so? This year’s temperature has dropped earlier than usual. Though only early November, it feels chilly. The butterflies I saw outside the window two months ago are nowhere to be found. Nonetheless, I know they will come again next June. Why am I so sure?

My window faces east. Since I do my writing in the evening, I have not seen the sun rise for several years. But I dare bet with anyone that if I sit by my desk in the morning and look outside, I can see the sun. When I see the ocean, I know seawater is salty, and that high tides and low tides are somewhat related to the waxing and waning of the moon. I was a fishing expert when I was a kid. When watching the sea, I could remember the happy moments of fishing. Anglers owe fish, but they understand the character of fish. A cloudy night with a full moon is the best time to catch sea bream. This is regularity.

Every learned man concurs to the laws of nature. Human behavior is no different. Overlooking from my window, Chi Fu Fa Yuen, a private housing estate built by Hongkong Land, sits right next to Wah Fu Estate, a public housing estate built by the Hong Kong government. No need for survey, everyone would agree that the latter is more densely populated than the former. Baguio Villa, a residential estate closer to where I live, has an even lower population density than Chi Fu Fa Yuen. The higher class the residential area, the lower the population density. This is regularity. On a nearby hillside, a few wooden huts scatter here and there. Illegally built, these huts are simple and plain. Illegally-built properties, having no land rights, are a lot simpler and plainer than properties with land rights. This is regularity, too.

Certainly, regularity can be traced irrespective of whether it is a natural or human phenomenon. In fact, it is impossible to find any phenomenon that exhibits no regularity – though certain phenomena require in-depth study to find their pattern. It has always been like this since ancient times that every phenomenon has regularity. We know it is like this, but do not necessarily know why. Since we know it is so, it is only natural that we would like to know why it is so. Curiosity is human nature. Since we would like to find explanation, science is thus derived.

Science is based on three key principles by which everyone interested in science must abide. First, the existence of any phenomenon or behavior is based on subjective judgment, and no deviation from this subjective judgment is allowed. When I say the sun is rising (my subjective judgment), if you do not agree and believe the sun is declining, then the two of us cannot get together to scientifically explain the sun’s phenomena. When I see flower, you see flower, too; when I say it is raining, you agree it is raining, too – This is the first prerequisite of scientific generalization. Of course, some people never agree to anything. These people will always be cut off from science.

Surprisingly, the agreement to subjective phenomenon is often easily achieved. Even if subjective judgments toward a certain phenomenon are different, it is still not difficult to agree on its existence. For instance, people with color-blindness would agree the colors that they cannot see exist; deaf people would agree that sound exists though they cannot hear.

It is one foundation of science that subjective phenomenon bears objective consent and common belief. However, certain subjective matter does not have the consent or common belief of the public, hence falling outside the scope of science. For instance, China in the old days boasted a lot about extra-sensory power. Believers believed firmly, yet there were lots of non-believers as well. I saw in Beijing the performance of the most-renowned extra-sensory power. It was so spurious that I never believed in it. Extra-sensory power lies outside the scope of science, not only because I do not believe, and not only because many people do not believe, but because it has never been rigorously tested to make non-believers believe. Just like some people believe in God while some others do not, yet no one has ever successfully confirmed the existence of God. This is not to say Christianity or other religion makes no sense, only that religion is no science.

The second principle of science, as aforementioned, is: every widely identified phenomenon is traceable and exhibits regularity. The regularities of certain phenomena may require massive effort to identify or confirm. Experience tells us that the regularities of phenomena have always been unchanged. Even though it is no easy matter to identify the regularity of a new phenomenon, scientific researchers, firmly believing in its existence, will never be daunted by repeated setbacks in its hunting.

Why is the regularity of a phenomenon so essential? The answer is: if a phenomenon only happens casually or at random without any regularity, there is no way to find out the relationship between this and other phenomena, rendering this phenomenon unable to be systematically explained. Non-traceable phenomenon shows neither signs in advance nor basis afterwards, similar to the Ascension of Jesus which cannot be inferred by logic. Science being science is because there are no phenomena in the world that exhibit no regularities.

This leads to the third necessary principle. Scientific researchers must firmly believe that nothing happens with no reason. To predict (not forecast) and to explain are the same thing. If we predict that under certain circumstances, for certain reason, certain phenomenon will happen, then the emergence of that phenomenon is counted as being explained. For instance, the speed of a fly is not as fast as that of an airplane, but due to Newton’s law of gravitation, a fly can nevertheless fly forward inside a cabin. The same theory is applied in explaining that a fly can fly forward inside a cabin, as well as in inferring that outside a cabin a fly cannot fly faster than an airplane. If the speeds of a fly and an airplane exhibit no regularities, or under different circumstances their speeds cannot be compared, then we have no way to explain the flying phenomena inside or outside a cabin. How then can we start exploring science?

To win the consent of the public, subjective phenomenon has to have regularity, and there must be a reason for it to occur or emerge. These are the necessary conditions of science.