Several years ago, an indisputable crackpot paper on the hep-th archive would be a rare and refreshing event. We would be entertained, enjoying the increased diversity that a particular dense had author brought to the community. However, it seems that the times have changed. Lots of stupid people have learned how to use the machinery on the web, they gained some technological self-confidence, and percolated the system.
These days, it is not unusual to see three or more full-fledged crackpot hep-th papers a day. These crackpots apparently seem to plan to take over the arXiv. To see an example, let's look at some of the papers that appeared yesterday.
Non-relativistic gravity: OK but unmotivated papers
Like every day, there are about three papers about the Hořava gravity. I am not counting those to the crackpot category. This is just a relatively meaningless fad that has activated many people who have never understood why the equivalence principle, the diffeomorphism symmetry, or the local Lorentz symmetry are so paramount in any modern theory of gravity.
There is no new evidence that they are less paramount than what was determined by the key insights of the 20th century science. However, sociology has changed for a while. Some people began to write papers ignoring these essential principles of modern physics, so many other people suddenly find it OK to expose their own ignorance, too.
But let's look at some other, less serious topics.
Kober's Ur objects
Martin Kober studies Ur objects, such as Pilsner Urquell (the beer). He is convinced that every Hilbert space must have a dimension that equals a power of two - because "quantum theory must be constitutive for human knowledge". He also believes that this qubit structure of all Hilbert spaces implies a flat Minkowski space or general relativity.
Needless to say, every single proposition in this long sequence is nonsensical or wrong. It would be enough for one of them to be wrong for the structure to collapse. But these crackpots always surpass the minimum required amount of nonsense at least by an order of magnitude.
The Hilbert spaces don't have to have dimensions that are powers of two. Large Hilbert spaces never have to have any exact tensor product factorization to smaller spaces, and they usually don't. The human knowledge isn't based purely on base-two qubits. Even if it were, properties of the human knowledge don't necessarily become constraints on the laws of Nature. Whatever the answers to all the previous sentences are, they have nothing to do with the Minkowski space or general relativity. To summarize, the paper is just a pile of nonsense, peppered by randomly permuted rudimentary formulae from field theory.
Bert Schroer vs crossing symmetry
Mr Bert Schroer believes that particle physics has been in the crisis for 50 years or so, approximately since its birth. String theory is the universal villain for him - but his papers always choose an additional villain. In the past, he has chosen e.g. path integrals of Richard Feynman as the ultimate symptom of moral decadence.
Mr Schroer has no idea how the path integrals actually work and why they work so well, so he has claimed that they break down even in simple contexts, not realizing that the only thing that has broken down is Mr Schroer's brain. This time, he chose another unlikely foe: the crossing symmetry. For Mr Schroer, this is a new symbol of the crisis.
In reality, crossing symmetry is one of the key consequences of relativistic quantum mechanics, or relativistic quantum field theory, if I use the more "special" language. The wave equation has solutions with both positive and negative energies. The negative-energy solutions are linked to antiparticles but because they arise from the same fields, their properties have to be connected to the properties of the ordinary, positive-energy particles.
See 70 minutes of a Feynman lecture explaining the reason for antiparticles.
This relationship is encoded into identities that involve the analytic continuation of various amplitudes to negative energies. Such a rule is general and it holds non-perturbatively. Antiparticles - and antimatter - are important and the crossing symmetry is a basic feature that defines them and constrains their properties.
Mr Schroer doesn't like it exactly because it's been so profound and important in physics of the last 50 years - and physics of the last 50 years is what he hates so much. He waves his hands, arguing that a modular localization of states shows some bug with the crossing property but his words make no sense. The coherence length scale, above which any rational relationship between the propositions and arguments evaporates exponentially, is comparable to 1 sentence in his text.
Incidentally, the crossing symmetry is an obvious part of string theory. In perturbative string theory, since its very birth, the crossing symmetry has been supplemented by a related but stronger principle, the worldsheet duality (that's the duality after which string theory used to be called "dual models" 40 years ago). This duality not only relates diagrams where external legs are interpreted both as particles and antiparticles, but it even relates tree diagrams with s-channel, t-channel, and u-channel exchange.
The evolution has been clear: as people have been unifying their understanding of physics, the quantities describing an increasingly diverse collection of processes were computed from an ever smaller number of independent diagrams. We insert less - but we obtain more. This is what the conceptual progress in theoretical physics is all about. Some of us are amazed by this property of Nature, others are devastated. ;-)
Thomas Larsson and jets
Finally, Thomas Larsson, a commenter on the blogosphere, discusses anomalies in QJT - which stands for Quantum Jet Theory. This is supposed to be an observer-dependent generalization of Quantum Field Theory that forces you to add additional terms of Taylor expansions to the quantum fields. Except that, as you might expect, such a generalization only exists in Mr Larsson's imagination. In this sense, QJT is fully analogous to Matti Pitkanen's TGD.
That can't prevent Mr Larsson from discussing anomalies and their dependence on the regulator in QJT. Throughout his text, he assumes that the reader believes that QJT makes sense and is, in fact, superior to QFT. The second physicist who believes that QJT makes any sense has yet to emerge. Larsson's propositions about QJT are analogous to similar basic statements about QFT, except that he randomly combines 50% of correct statements with 50% of wrong statements.
Let me summarize. I used to say that the non-peer-reviewed character of the arXiv didn't hurt. Clearly, this bold proclamation reflected the actual reality that existed at that time and that reflected the actual community of people who were able, allowed, and willing to submit papers to the arXiv. There has never been any guarantee that such a filtering procedure would remain forever unnecessary. In fact, as the concentration of pure crap is increasing on the arXiv, the need for a more constraining filtering procedure is gradually becoming urgent.