If "this raven is black" is true, it has nothing to do with the specificity of the statement, but because the statement can be proven by a universally repeatable controlled observation. So, can universally repeatable controlled observations prove that "all ravens are black"?
In this issue, the distinction between "observable" and "controllable" is key. The ravens in the above statement are only "observable variables" in controlled processes, not "controllable variables"; Proving that "this raven is black" is true is a "controlled observation", not a "controlled experiment". The difference between controlled observation and controlled experiment is only that the former can only control the observation conditions, while the latter can not only observe the object again and again, but also make the observed object or some properties of the object controllable change by changing the controllable variables.
The basis of science is controlled observation and controlled experiment, and the so-called objective reality is only based on the fact that controlled observation is generally reproducible, which is the truth emphasized by science before the formation of modern science. What we call the authenticity of scientific experience is the universal reproducibility of controlled experiments and controlled observations.
This article is the third in a series of five articles in a series of Chinese studies on the proposition "Philosophy of Science in the Shadow of the 'Linguistic Turn'". The original article was excerpted from Mr. Jin's new book "The Disappearing Truth", and today's theme is "Can't the Full Name Statement Be Confirmable".
Isn't the full name statement verifiable?
Text/Jin Guantao
Since being confirmed has nothing to do with the specificity of the statement, can the full statement be confirmed? As mentioned earlier, "This raven is black" is true not because the black raven exists objectively, but because the statement can be proven by a universally repeatable controlled observation. So why doesn't universally repeatable controlled observation prove that all ravens are black? "All ravens are black" is speculation because the ravens here are only "observable variables" in controlled processes, not "controllable variables". What is a "raven"? In animal classification, a bird called a "raven" can be defined by a set of attributes: they have shape t1, feeding t2, anatomical characteristic t3, etc. where t1, t2, t3... These are "observables", not "controllable variables".
A raven is actually a collection of observable variables. The statement "all ravens are black" expresses the following observation: we observe characteristics t1, t2, t3... The n birds of the set (note : black does not belong to the set) are black, so we made the statement "all ravens are black". Obviously, I can't guarantee that the n+1st raven observed must also have black characteristics, so "all ravens are black" is a guess. But when it comes to dictating that Raven is a set of "controllable variables," the situation is quite different.
Define the trait set of ravens t1, t2, t3... Corresponds to a group of genes τ1, τ2, τ3... As long as you control a set of genes, you can make one with t1, t2, t3 in the laboratory... Trait of birds. If they find out one day, with the control genes τ1, τ2, τ3... The birds produced by the method are black, and this experiment can be universally repeated, that is, as long as any experimenter achieves this set of conditions, he must be able to synthesize a trait set t1, t2, t3... of birds, and this bird must be black. Biologists are not surprised by this, this is because of the genes τ1, τ2, τ3... Not only the traits t1, t2, t3 of the bird are specified ... And also prescribe its color (note that the experiment here is imaginary, which is not necessarily the case). At this point, can we say that "all ravens are black" is just a guess? No! Now it has been confirmed as well as "this raven is black". In other words, if we admit that "this raven is black" can be verified, we must also admit that the aforementioned "all (man-made) ravens are black" can also be confirmed.
Here, attention must be paid to two premises. First, it is necessary to distinguish between which variables are controllable and which are merely observable. In the full-name statement "all ravens are black" as speculation, t1, t2, t3 ... are observable variables, and in the corresponding corroborated universal statement, τ1, τ2, τ3... are controllable variables. The former defines ravens only by observation, while the latter creates ravens by controlling activities. Why do I dare to assert that the second universal statement will not be falsified by the discovery of black swans, as "all swans are white"? In the gene collection of the prescribed swan, the gene that specifies the color and the genome that specifies the swan morphology can be separated.
Removing a certain gene causes a change in swan color. In other words, the genes that prescribe the genome of the swan and the color of the bird collide by chance, and once the color gene is mutated, a black swan immediately appears. This is unlikely to happen for the aforementioned artificial ravens. The reason is simple, the above experiment is controlled and can be repeated arbitrarily. The experiment stipulates the color of the biological trait while controlling it. At this point, the possibility of having a prescribed color gene is completely ruled out. In this way, in the real world, there is no longer a sudden change in the color of the raven. If a new species that is not black appears, the bird's traits must change, and it is no longer a raven. The key to confirming whether this link must exist (color is controlled by trait genes) is whether I can control the "raven" variable, that is, by controlling the genome to make birds with the trait "raven" (including knowing whether a gene can be removed to change the color of birds with raven traits), otherwise I will never know whether the property of black is dictated by the raven trait. When the raven trait is a completely controllable variable, the reproducibility of the experiment ensures that "all ravens are black" can be confirmed.
The second premise goes to a deeper level. Why as long as experimenter A achieves conditions τ1, τ2, τ3... When phenomenon E is observed, can another experimenter B observe phenomenon E when the same conditions are achieved? Many people think that this may not be possible, so the corresponding full name statement is just a guess. Even if 10,000 observers repeat the experiment of water freezing at 0°C and one atmosphere, we cannot guarantee that 10,000 observers will observe the same phenomenon. So, can we say that the blanket statement "water freezes at 0°C and one atmosphere" has been confirmed? Of course not. In this sense, the Hume problem has not yet been resolved. The fact that the first n observer repeats the experiment does not mean that the n+1st observer can also repeat the experiment.
Please don't get me wrong, what I want to prove is that if the conditions for recursive confirmation are not true, a single statement cannot be confirmed. After a raven has been shown to n people and identified as black, we also cannot guarantee that when shown to the n+1st person, the raven's color will not change so that it will be black to new observers. Because for the new observer, time, place, or other conditions are always different from the original observer, even for the same observer, the "self" of the next moment is still the same as the previous moment.
That is, if a single statement must be verifiable, then the recursive confirmation I proposed must be true, and the propositions proposed above that are regarded as speculation by falsification must also be verifiable. In fact, this problem also exists for the "self". If the "self" is thought to exist, a recursive confirmation procedure is also required. Therefore, my equating the full term with recursive enumerable assumes that the self and free will exist, that authenticity is only an extension of it, and behind it lies the uniformity of time and space, and the stability of self-consciousness (i.e., the observer still thinks that he is the same as the previous moment at the next moment). These more in-depth questions cannot be discussed in detail here, but I will discuss them in the methodology and construction chapters.
In fact, what statement confirmed by controlled experiments is not universal? Is "pure water boiling at 100°C and one atmosphere" a guess? No. Is it a single statement? Nor is it. Are there near-infinite details such as the time, place, and observer of the experiment in statements such as "pure water", "temperature 100°C", and "one atmosphere"? No. They are controlled variables, and experimental reproducibility requires them to be repeated again and again. Any condition statement actually covers the full name of this condition, behind which is the repeatability of the control condition, and observing the phenomenon of water boiling is a process of recursive confirmation. Here, not only does the controlled variable contain the invisible full term, but the object "water boiling" is also a full term. As we all know, most of the scientific facts we know are some substantiated full-term statements. Scientific theories are based on universal statements, and they cannot be mere falsifiable speculations.
The difference between controlled observation and controlled experiment is only that the former can only control the observation conditions, while the latter can not only observe the object again and again, but also make the observed object or some properties of the object controllable change by changing the controllable variables. At this time, in addition to the observation conditions, the control also includes a certain property of the object. When the above control processes are generally repeatable, the controllable nature of the object is also a controllable variable. We can add it to the set of controllable variables of the initial experiment, constituting a self-iteration of the controlled experiment to form a new controlled experiment.
It can be seen that controlled observation is only a special kind of controlled experiment, that is, those controlled experiments that cannot be self-iterative. I will discuss the basic structure of controlled observation and controlled experiments, and under what conditions a controlled experiment transforms into controlled observation. Here I just emphasize that the basis of science is controlled observation and controlled experiment, and the so-called objective reality is only based on the fact that controlled observation is generally repeatable, which is the truth emphasized by science before the formation of modern science. What we call the authenticity of scientific experience is the universal reproducibility of controlled experiments and controlled observations.
This article is excerpted from Chapter 3, Section 3 of The Vanishing Truth. The comments have been deleted, please refer to the original text for academic discussions. The picture comes from the Internet, if there is any infringement, please contact to delete.