Wikipedia talk:Actually editing scientific articles

Why this page is rubbish

This page is written by people who are ignoring the fact that scientific articles are never written or verified according to the simpleminded "check the sources" method. Many sources, especially textbooks, say wrong stupid things when it comes to science. This is true in other places too, but in fields such as physics, or mathematics, the answer is objective, so it is possible to do much better than to say "X says A but Y says B". It is also possible to say "according to the consensus understanding, X says A because A holds under these circumstances, and Y is only saying B because B is valid under these different circumstances".

This type of thing happens all the time. For example, every physicist knows what "E=mc2" means. But sources often state it in a misleading way. They say that "mass can be converted into energy". This makes it seem that when mass is converted into energy, the system weighs less. So that when an atom bomb explodes, releasing energy, the total mass of the bomb goes down. That's not true, because the heat and light released weigh just as much as the missing mass, so the total mass-on-a-scale is constant. But sources that make the statement "mass can be converted into energy" hardly ever go into details about what they mean, nor do they discuss this example. This is one of many trivial conclusions that anyone who unerstands relativity comes to on their own, and which can only be checked by thinking and understanding.

Consensus understanding is defined by reading the sources with understanding. It means that when a source says that "antiparticles are particles going back in time" and another source says "antiparticles used to be thought of as particles going back in time, but this is no longer considered a valuable point of view" (this was a recent dispute), it is possible to resolve this dispute by studying the formalism of quantum field theory. Then you can see whether the back-in-time formulation is still used by physicists, or whether it is superseded. Similarly for "positrons are holes in the vacuum" picture. This is also claimed to be historical, like Maxwell's gear wheels, and to some extent it is. But it is much better to explain the replacement of creation and annihilation operators which converted Dirac's hole picture to the modern quantum-field picture. This replacement has a source, but it's a technical source from the 1930s, and it does not directly adress the controversy. So when summarizing this source, one needs to actually understand what it says.

An example from the WP:ESCA talk page, which came up on "thermodynamics". Some sources say "dU= Tds -PdV", other sources say "dU< TdS - PdV". Which is it? Everyone who knows thermodynamics understands the two formulas: the first applies to adiabatic changes, the second to cases where entropy can go up. But the books usually don't say this explicitly, leaving it implicit by context. This must be sorted out by explaining it from first principles, otherwise the debate can go on forever.

Further, the best way to write this formula is "dS > dU/T + P/T dV", so that the entropy quantity is on the left. This replacement is found in some sources, and recommended as a style of writing thermodynamics in some sources, and it is clearly equivalent to the original mathematically. But it is easier to relate to statistical mechanics. Further, the quantity "T" is better replaced by "beta=1/T", and Boltzmann's constant is better set equal to 1. These stylistic issues are not going to get settled by sources, and if they are not done just right, the formulas become incomprehensible.

With good faith discussions from first principles, it is very easy to find the proper phrasing for scientific articles in almost all cases. The few remaining cases are those where there is actual confusion in the professional literature, and those confusions must be represented with NPOV just as all other disputes. But to put statements culled from ignorant textbooks side by side with statements culled from the professional literature is crazy. It fills the encyclopedia up with nonsense, and creates disputes between editors when there is no real dispute in the professional literature.

It does not require a lot of expertise to do any of this. Anyone can read most of the sources, and the technical sources can be read with a little help by anyone as well. If an editor has difficulty with the technical material after patient study, and patient explanation, then perhaps this editor should study it further before editing the article. Without this process, the articles become a hodgepodge of conflicting claims by different authors of no value to the reader, since, unfortunately, actual scientific knowledge does not come in soundbites copied from textbooks.

The most egregious cases are the presentation of mathematical proofs. Mathematical proofs are the most difficult text to write. A mathematical proof of something as simple as 1+1=2 can be made incomprehensible (see Principia Mathematica for an example), and the most difficult proofs in the literature can be made transparent. It all depends on the language.

The way in which to write proofs is defined by thinking. You keep the ideas, but you change the formulas and the language until it becomes as clear as possible. Every author does this in a different way, but the ultimate standard of success is clarity and accuracy. Most textbooks present proofs in the worst possible style.

My personal interest in this matter: I wrote a proof for the page Godel's incompleteness theorems, and the proofs were rejected because they are not found verbatim in sources. This despite the fact that the proofs are paraphrases of existing proofs, and despite the fact that they are super-clear. There is no new idea in the proof, but it is stylistically different. The arguments that I got rejecting the proofs led me to believe that Wikipedia's policies regarding scientific articles are not optimal.Likebox (talk) 21:04, 25 January 2010 (UTC)