Could it be that a TOE cannot be scientific?

[Ivan Seeking]

One interesting notion that strikes me from time to time, like now, is that if String Theory [ M Theory] is correct or nearly so, the greatly anticipated theory of everything, or TOE, is not a theory of science and never can be. Rather, M theory stands as a philosophical proposition, and it appears tha tthis will always be true. As far as we know, M Theory can never be tested or qualify as a theory of science. So, this makes me wonder if this might be implicitly required somehow; perhaps due to a philosophical limitation built into science that we've missed? Anyway, not to be taken too seriously, it's just a thought that I have now and again.

 

[Pi_314B]

The TOE will be scientific, but it will be so simple as to make you think that it is not.

 

[metacristi]

One interesting notion that strikes me from time to time, like now, is that if String Theory [ M Theory] is correct or nearly so, the greatly anticipated theory of everything, or TOE, is not a theory of science and never can be. Rather, M theory stands as a philosophical proposition, and it appears tha tthis will always be true. As far as we know, M Theory can never be tested or qualify as a theory of science. So, this makes me wonder if this might be implicitly required somehow; perhaps due to a philosophical limitation built into science that we've missed? Anyway, not to be taken too seriously, it's just a thought that I have now and again.

Yes it seems that we have a problem with falsificability here (Popper's version), indeed an unfalsifiable theory could, still, be at least approximately true. But it's too strong to say that superstring theories cannot be tested, this remain to be seen (the theoretical work is still in progress and the hope is that at high energies they could, potentially, be 'confirmed').

In fact we have even now a clear novel prediction of string theories which is considered as potentially testable in the very near future (when the european collider will be over) namely supersymmetry. The problem is that many other theories do predict the existence of such super partners so this is not regarded as a crucial step ahead into accepting string theories as confirmed. Anyway, as enthusiasts point out well, the corroboration of supersymmetry would boost our confidence in them, that we are on the 'right track', at least that it is worth pursuing further this path as the first choice research program (this is especially important if we think at the funding of new experimental devices, very expensive, needed to, potentially, find crucial 'confirmations' of the theory).

What really put in doubt the scientific status of string theories is the fact that, apparently, they have adjustable parameters enabling them to accommodate all possible experiments, being thus basically irrefutable. For example if a version of the theory is falsified by a certain experiment (at very high energies for example) there is always room, critics say, for 'post hoc' (post fact) changes to accommodate the anomaly by adjusting those parameters.

If so then string theories are more or less in the same category with Adler's theory or Freud psychoanalysis, two well known examples of pseudo sciences indicated by Popper. However the popperians can retort that such an approach is acceptable as much as the revised version of the theory not only accommodates the existing anomalies but makes also novel predictions (potentially testable) different from those made by the earlier version of the theory (psychoanalysis or astrology for example only accommodate facts 'post hoc' without making novel, potentially testable, predictions from their new accepted premises). Remain to be seen if string theories will be able to withstand these requirements, they may why not, as of now at least there is nothing we are aware of which to impede this.

Even if the different variants would be superseded very often if the paradigm change takes place as presented above then we could still say that string theories are scientfic upon Popperian falsificationist methodology. Only the absence of novel, different predictions, could justify scientists to label it, potentially, unscientific.

But this in no way totally discredit such theoretically and experimentally stagnant programmes, and this is a real weakness of falsificationism in Popper's variant. Though they look stagnant at a certain moment such programmes could still become become progressive in the future, the history of science does offer us such examples (even seemingly degenerative programmes at a certain moment became progressive later).

Happily the mere popperian falsificationism can be improved, Lakatos presented an improved version of falsificationism which takes into account these problems, only if a program is degenerative or stagnant for very long periods of time is it totally abandoned, being considered 'falsified' (in a weak sense of course, scientists are always prepared to realize later that it is still viable). So I'd argue that string theories are scientific even now, the mere Popperian account has proved unable to make a clear distinction between science and pseudo science, its principal aim. This by no means amount to say that popperianism has to be discared, it is still a very useful methodology in the vast majority of cases, but we must be aware of its clear limitations.

 

[metacristi]

Few clarifications: I said above that string theories are scientific even now. This is because the supersymmetry is a novel prediction, so even in the popperian methodology (though only at limit) they are at least scientific hypotheses.

Secondly they can be seen, at limit, as being theoretically progressive (being able to accommodate an increasingly greater number of already known facts). Thirdly they involve simple and unified ideas, one of the aims of the actual acception of rationality. Fourthly they are totally compatible with the Standard Model and can be seen as a natural extension of it, with only minor exceptions (this 'coherentist' view is not very strong but anyway I'd argue that a form of reliabilism can be accepted at least at the beginning of the quest for a new paradigm).

Fifthly I'd argue that superstring theories can be designated currently (orientatively) as the first choice research program, they not only accommodate an increasing number of already known facts but in the same time they are more fruitful currently (in what the number of facts accommodated is concerned) than loop quantum gravity.

There is no good reason to think that the situation will stay the same in the future but as the situation presents today it is reasonable to accept this, provisionally of course (thus it deserves to pursue alternative programmes too, the most promising amongst the existing ones being, in the light of the actual acception of rationality, loop quantum gravity).

 

[selfAdjoint]

If any theory were truly confirmed as a TOE, it would have to be demonstrated that this was so. It is conceivable that this demonstration could be logical or mathematical, and if mathematical perhaps non-constructive. That is you could assume that the given theory did NOT explain "everything", in whatever definition of that word was being used, and then derive a contradiction rigorously from that assumption. If the excluded middle is accepted then that would show the theory DID explain averything without in any way describing HOW it does that. To make things even worse imagine the proof involves the axiom of choice.

I predict that if this occurs we will get panjandarums pontificating that if the public is uncomfortable with that outcome it's just because the public is stupid and ignorant.

 

[kishtik]

God created all. This is the Theory of Everything and it is not scientific. It answers all of the questions you can ask. `Because God wanted that way...`. So apart from the lack of curiosity of the deeply religious people (not offending), if there will be a TOE, it should be scientific because we have a non-scientific one and we don't need another.
I hope that our experimental devices will one day be stronger than the creativity of our theorists. That was true before Newton

 

[nameless]

I have found a reasonable ToE. It derives from a synthesis of quantum mechanics, consciousness (which is the ingredient that has been glaringly missing from hard science, but not for much longer...) studies, mysticism, metaphysics, various personal disciplines, etc...
It is pretty simple, but as it is a 'synthesis of all relevant disciplines' it is not testable or verifiable completely by anyone 'hard' system. On the other hand, all it violates is our antiquarian belief structures which we hold so dear and emotionally are loathe to relinquish, while all 'hard' questions are answered or dismissed by the ToE. There is so much emotional, egoic and religiously dearly held and defended ignorance, that a new and simple ToE is as nothing in in the face of the human hurricane of ignorance and stupidity. For example, how long did it take the 'public' to accept that the Earth is not the center of the universe even after the 'scientists' showed the evidence? Hasn't Gallileo just recently been 're-communicated'? *__-

I hope that our experimental devices will one day be stronger than the creativity of our theorists. That was true before Newton

Hi Kishtik, I'm afraid that you have it reversed. Once the theorists 'create' something to 'find', only then can the 'technicians' develope the technology to actually 'discover' what the 'theorists' have... 'created'. It has always been such. First something must 'exist' in the 'mental realm' before becoming 'apparent' in the 'physical'.

 

[Locrian]

In fact we have even now a clear novel prediction of string theories which is considered as potentially testable in the very near future (when the european collider will be over) namely supersymmetry.

Supersymmetry is absolutely not a novel prediction of string theories, nor could its testing be considered a test of string theories.

 

[metacristi]

Supersymmetry is absolutely not a novel prediction of string theories, nor could its testing be considered a test of string theories.

Superstring theories, at least the current versions, definitely incorporate supersymmetry counting as a novel prediction. The real problem, as I've explained also above, is that its possible corroboration ('confirmation' without assesing also that the theory is more probable to be true) cannot be considered a crucial experiment. Which is totally another thing. Anyways there is no sufficient reason now to say that superstring theories are not scientific. Even in the mere Popperian account. Let's better see what the future can bring to us...

 

[selfAdjoint]

Superstring theories, at least the current versions, definitely incorporate supersymmetry counting as a novel prediction. The real problem, as I've explained also above, is that its possible corroboration ('confirmation' without assesing also that the theory is more probable to be true) cannot be considered a crucial experiment. Which is totally another thing. Anyways there is no sufficient reason now to say that superstring theories are not scientific. Even in the mere Popperian account. Let's better see what the future can bring to us...

Superstring theories imply supersymmetry, but so do non-superstring theories like GUT and the MSSM (minimal supersymmetric model), both non-stringy extensions of the standard model with supersymmetry. The new LHC collider at CERN is expected to be able to say yea or nay on supersymmetry; the nay would falsify superstrings, but the yea would not be evidence for it, according to standard logic.

There are versions of string physics that do not use supersymmetry; bosonic string theory is one, and there was an early version of superstring fermionic theory that didn't use full-bore supersymmetry but used Grassmann variables in expressing the length of the string (I learned this from Krauss's new book Hiding in the Mirror). And then there's string field theory which is a whole new ball park.

 

[metacristi]

Superstring theories imply supersymmetry, but so do non-superstring theories like GUT and the MSSM (minimal supersymmetric model), both non-stringy extensions of the standard model with supersymmetry. The new LHC collider at CERN is expected to be able to say yea or nay on supersymmetry; the nay would falsify superstrings, but the yea would not be evidence for it, according to standard logic.
;

If one reads carefully my previous posts here one would easily realize that I acknowledged that the possible corroboration of supersymmetry in experiments at very high energies cannot count as a crucial experiment. However, strictly from a logical standpoint, supersymmetry is part of superstring theories, thus it does count as a novel prediction of superstring theories (in other words supersymmetry does count as a novel prediction for all theories predicting / incorporating it; it is 'novel' not because other alternative theories [do not] make it but because the previous paradigm, considered as 'normal science', did not have it).

A crucial experiment, is one whose interpretation of 'bon sens', to quote Duhem, makes the majority of scientists to accept a scientific hypothesis as being the 'normal science' of the time (as was for example Rutherford's experiment for the atomic hypothesis or the experiment corroborating Relativity in 1919).

There is no need for a theory to make a lot of novel predictions, many corroborated, to become 'normal science', a single corroborated novel prediction, if considered crucial, is sometimes enough (of course with the condition that other predictions of the theory are not falsified and it should prove also progressive, at least theoretically, in the medium run).

Sure in the most cases it is the fact that theories prove very progressive (they ha[ve] many novel predictions corroborated + many already known facts accommodated / unified) which make the scientific community to accept them as 'normal science' but this is not always the case, as I've stressed above.

It's true now that in the philosophy of science there is plenty of room to doubt that there really exist crucial experiments in physics at least (which entails that physics surely approach Truth, in absolute, not at all granted from all we know now) but it is still rational to accept the interpretation of some experiments as being 'crucial' as much as fallibilism is never dropped (some of my thoughts related with this topic, whether there are really crucial experiments in physics, are here).

There are versions of string physics that do not use supersymmetry bosonic string theory is one, and there was an early version of superstring fermionic theory that didn't use full-bore supersymmetry but used Grassmann variables in expressing the length of the string (I learned this from Krauss's new book Hiding in the Mirror). And then there's string field theory which is a whole new ball park.

I talked here only of superstring theories, if I understood well the OP talked rather of M-theory, attempting the unification of the five types known of superstring theories ('super' comes from supersymmetry :-) ), seemingly the most promising research program currently. No serious thinker would argue that if supersymmetry is falsified then string theories are a 'dead end'. By the way Krauss view is highly controversial, apparently he is incapable to make the distinction between potentially testable (that is enough clear, potentially testable, predictions exist-this is the case of superstring theories) and basically no potential testability at all (the case of ID currently, of course this in no way implies that the situation will ever remain so, but this is how the situation presents now).