Everyone knows by now about the LIGO announcement regarding the detection of gravitational waves. However, some things must be clarified about the relation of the detection to claims of a test of the theory of general relativity and about the way the media has presented this to the general public.

I recently attended a presentation about the LIGO detection given by Dr. Apostolatos of the University of Athens. Dr. Apostolatos earned his PhD under the supervision of Kip Thorne, who is a physicist at Caltech and who has played a leading role in LIGO’s development. Therefore, Dr. Apostolatos was more than qualified to give this presentation. At the same time, besides being an expert in this area, Dr. Apostolatos is a charismatic speaker and he was able to clarify several issues in ways that the general public can understand them. It was an excellent presentation, one of the best I have ever attended on a physics subject.

To start with, I will assume that LIGO detected a gravitational wave, despite the fact that this type of detection appears extraordinary and according to some even impossible. Remember that LIGO claimed to have measured a wave that forced the planet Earth to expand and contract by 1/100,000 of a nanometer. Is this type of measurement possible given the known and unknown sources of noise and detector limitations? The claim is that it is and I will not dispute that. I will just assume that what was measured was indeed a signal of a gravitational wave.

Therefore, the following is true:

Conclusion: A gravitation wave was detected

The truth of the conclusion rests primarily on the following:

  • The form of the signal is what we expect to get from a gravitational wave
  • The signals was detected by two different detectors with the expected lag
  • All potential sources of noise were isolated

As I said, I will assume that there are no problems with the above and the signal was from a gravitational wave. Thus, I accept the truth of the conclusion.

However, here is where the ambiguity starts: For example, in the public media it was reported that due to the merger of two black holes, a gravitational wave was generated that was measured by LIGO. Since this type of a wave is predicted by general relativity, this test vindicates the theory.

But there is a problem with this reasoning. Remember that black holes are not directly observable due to their nature. Actually, what was shown in the public media, i.e., the two black holes merging and the blast that followed, was a computer simulation of a possible fact using the equations of general relativity. General relativity is the theory that the gravitational wave detection intends to provide support for as the second part of this experimental investigation. Thus, the actual fact that generated the gravitational waves was abducted in a logical sense and was not arrived at either by deduction or by induction.

Fact: There was a merge of black holes that generated gravitational waves

The rule by which the fact was abducted was the model of general relativity:

Rule: The model of general relativity

Why is this problem worth discussing and taking the reader’s time? It is important in my opinion because the use of logical abduction may be problematic in physics and may lead to some dark paths.

Let me regress for a moment and attempt to explain abduction. You all have heard of deduction, a logical process via which a conclusion is made from a rule and a fact.  An example of deduction is as follows:

Fact: Wilhelm came out of that house
Rule: All residents of that house are German
——————————————————-
Conclusion: Wilhelm is German

In deduction, we know that a rule is true, we have a true fact and we reach a true conclusion. If the rule or the fact or both are false, the process of reaching the conclusion is valid but not sound. Therefore, in deduction we have to make sure that the fact and the rule are true. Obviously, this was not the case with the LIGO detection of gravitation waves because we first assumed the conclusion to be true, then we derived (abducted) the fact using the rule and then assumed that the rule is true. Well, some may try to argue that the conclusion corroborates the rule but that is more like a word play.

In induction, we have a true conclusion and a true fact and from those we try to derive a general rule, as shown below:

Conclusion: Wilhelm is German
Fact: Wilhelm came out of that house
——————————————————-
Rule: All residents of that house are German

I am not going to get into the well-known philosophical problem of induction but for the above inference to be sound, the conclusion and fact must be both true. However, this was not the case with the LIGO detection because the fact was assumed true after assuming the rule true.

In abduction, we attempt to infer a fact from a true rule and a true conclusion:

Rule: All residents of that house are German
Conclusion: Wilhelm is German
——————————————————-
Fact: Wilhelm came out of that house

As relating to the LIGO detection, I already assumed that the conclusion was true and the signal was actually from a gravitational wave. But in order to establish the soundness of the fact, i.e. that this occurred because two black holes merged, one must assume that the rule is true, i.e., the theory of general relativity is correct, as follows:

Rule: The model of general relativity
Conclusion: A gravitation wave was detected
——————————————————————————————–
Fact: There was a merge of black holes that generated gravitational waves

Therefore, there is certain circularity with the above abductive inference. The rule is neither an axiom, nor established by induction; it is actually what we like to test in terms of a theory. Then, the fact, i.e., that there was a merger of two black holes, was derived after we applied our conclusion to the rule. Therefore, even if the conclusion is assumed true, little can be said about the rule.

In the CERN experiment to detect the Higgs boson, the fact is caused by the collision of particle beams at high energies.  Thus, the fact is known and the detection is based on collecting data from many collisions, i.e., it is an induction, not an abduction. Specifically, from the statistical analysis of many signals (the conclusion) resulting from collisions (the facts), a rule is induced, i.e., that there is an excitation of the Higgs field and explains why some particles have mass although due to symmetry they should have none.

When I asked Dr. Apostolatos a question after his talk about the use of abduction and the difference from the CERN experiment, he replied that also in this latter case we do not directly see the Higgs particle but also some detector signal. However, there is a big difference, as I hope I explained above. In the CERN experiment we do not abduct the particle collisions; we create them. Hence, we can establish causality, even in a countefactual sense. But in the LIGO detection we assume that the experiment is taking place about a billion light years away and has a particular form: two black holes merging, because this is what the theory predicts. Therefore, in CERN we had an experiment with a generated cause and effect but in the LIGO detection we have only an effect and we abduct the cause of it, i.e., the black hole merger, through the model of general relativity.

The job of a philosopher is to raise a question. In this case the question is whether abduction is compatible with the scientific method. The job of the scientist is to try to provide an answer in a way that fundamental and indisputable rules of philosophy are not violated. Science needs philosophy to evaluate epistemological and ontological claims and philosophy needs science to ask the proper questions that do not violate empirical truths. I do not know the answer to the question. Maybe it is trivial but maybe it is not. Maybe in the longer-term it will not matter. Maybe it will matter. I stop at the point of asking the question. After all, someone once said that “the question is, what the question is.”