Bob_for_short

So what is the question? Of course, an atom recoils while radiating.

Absolutely correct. Again, if you have at least two interacting particles, there are relative distance r = r1 - r2 and the interaction potential depends on it, and the center of inertia coordinate R. Equations for R and r are separated. Equations for r may describe bound states in case of attractive potential and even reactions (a complex potential with an imaginary part). So the interaction is the main thing - it determines the overlapping, if any.

There are two de Broglie waves - one describing the relative electron-nucleus motion in an atom (atomic size) and the center of inertia de Broglie wave (position of the atom as a whole). It is the latter that determines the interference or diffraction features, and it depends on the preparation device (source).

A compound system is described with its center of inertia variables, say, R, and relative motion variables, say, r. The wave function of relative motion depends on r. It determines the atom size. The center of inertia coordinate R determines the CI position in space. In QM the total wave function factorizes, so the CI wave function may be a plane wave or a wave packet, depending on preparation device. In case of a plane wave you may obtain an interference picture but it does not mean that an atom "balloons".

Yes, there is a connection called interaction. When two charges scatter (= interaction of charges), they produce photons and final charge states. In a particular case of charge annihilation the photon energy is just higher. It is just like atom-atomic scattering without and with nucleus fusion: the relative motion of atomic electrons in final atoms changes (gets excited).

I have a theory where an electron is a part of photon oscillators. Such is their intrinsic coupling in my construction. What is interesting - the electron oscillates in such a construction and the state of particular oscillators determines the total picture. It may look as "hidden variable" influence although in my construction the nature of variables is apparent and physical.

Yes, I think so, except I do not like the term collapse. Any observation is getting a specific information. Another matter that one-time measurement may not contain all information about the subject in question (wave function). Even in classical statistics every single observation has only one issue whereas many observations give the issue distribution (probability density). We do not call a one-time issue of dice experiment a "probability collapse". The notion of probability belongs to an ensemble, it is its characteristic.

Yes in principle and no in reality. To obtain entanglement one has to use the conservation laws. For that the initial atomic state and that of a projectile have to be well determined (by the preparation devices). After scattering the projectile energy loss should be measured very accurately to judge in what energetic state is the target atom in the final state. Resolving the projectile energy is a very hard problem. So in practice it is not done. In practice it may be easier to observe the target atom state directly (traces of ion and electron) or indirectly (atomic radiation).

If you scatter a projectile from an atom (a bound relative motion state), then the final atomic state is a superposition of all excited atomic states allowed by the energy conservation law. These excited states may include the ionized states (free relative motion of atomic components).

Our problem is in classical perception of psi-squared and trajectories. No, it is not a particle density. It is the probability density and to get it we need many experiments. For example, to draw the experimental atomic form-factor squared |F(q)|2 we have carry out many experiments. A narrow wave packet is still a wave function rather than a a localized particle. Why can't we look at the psi squared as at a photo of a complicated source instead of one electron/photon?

Let us consider two bodies connected with a spring. It is a mechanical system and is described with mechanical equations. I analyzed it (with another purpose though) and I see that according to the mechanical equations for, say, particle 1, it feels the force (attractive or repulsive depending on phase of oscillations) from the second particle. So if you hit the first particle quickly, its back reaction (effective mass) will essentially depend on the phase of oscillation. In other words, the spring (the glue) may "help" the external force push particle 1 or may "resist". We must keep this peculiarity in mind while analyzing the scattering data. The picture with static quarks and gluons is not realistic - they oscillate.

And if a particle of a finite size is not rigid but soft? Who said that it should be rigid? Who said that it should be elementary if it is in interaction? How to "detach" a charge from its electromagnetic field which is the main characteristic of the charge? Nohow. They are permanently coupled. They possess an infinite number of degrees of freedom. Some of these degrees describe the center of inertia of the system (3 coordinates suffice), the others describe "internal" or "relative" motion in the system (photons). I have a pet theory about this, if you like. Yes and no. Yes, because the point-like and decoupled electron is used as the initial approximation, and no because after renormalizations and the infrared problem resolving QED deals with the dressed electrons, i.e., with real electrons permanently coupled to the EMF degrees of freedom. My pet theory deals with the dressed electrons from the very beginning. According to my model, the real QED electron is "large" and "soft". You cannot push it without making radiation = perturbation of the initial relative motion state = radiation of photons. Cool?

To be a black hole, it is not necessary to be a Bohmian particle but a compact mass with the gravitational radius smaller than the body size, it seems to me. Next, a point-like "particle" is an inclusive (average) picture when you replace the real body with only three coordinates of its center of mass (or its geometrical center). All bodies are of finite size, actually.

I, as the author of a reformulation approach, am interested in a constructive feedback from researchers capable of understanding the issues in question and interested in resolving the old physical problems in electrodynamics. Feel free to discuss and contribute to developing this specific direction at http://groups.google.com/group/qed-reformulation Regards, Vladimir Kalitvianski.

In a non-relativistic approach every path is possible. Factually such an integral is a good zero anyway.

Yes, that's true. Many Worlds, Many Gods, Hidden variables, etc., but it is not serious in my eyes. In Classical Mechanics we have three coordinates with determinism but to observe these three coordinates in practice one exchanges with energy (photons) with internal degrees of freedom of a body. So the body is complex, not point-like. The internal degrees of freedom (relative motion of pieces of a body) is described with a plenty of other equations. They permit to get the right information about the center of inertia (3 coordinates) and sizes of the body. In QM, instead of additional set of equations, we have a wave equation covering "deviations" from the average (simplified, point-like) picture. This is much similar to those "internal motion equations" in CM. In both cases we need a lot of information to describe a phenomenon. This understanding saves us from searching for "wave function collapse" reason, hidden determinism, etc.

This is not "a valid interpretation" but a fact.

Thanks, Ajb, I thought everybody was aware of this but I could be wrong.

I do not hide my ideas and solutions. In my profile there is a reference to my web log. The electron is smeared quantum mechanically, of course. You know about negative charge clouds in atoms. The same thing for an electron coupled to the quantized EMF. The idea is very simple and physical: if pushing an electron excites a photon oscillator, then the electron is a part of this oscillator. Therefore it is smeared quantum mechanically and this is described with the oscillator wave function. In case of many photon oscillators the situation is similar, just like in many-electron atoms. For more details please read first my "Atom as a 'Dressed' Nucleus" article, published in CEJP and available on arXiv.

I used to work with supersymmetry and supergravity. At that time I also was exited with the beauty of ideas but frankly, supersymmetry follows from nowhere. It is not a phenomenological approach but mathematical. So it does not correspond to reality. The ideas I develop are purely phenomenological and do not need extra things. I just model an exact coupling of electron with EMF degrees of freedom. In exact coupling the charge is smeared, not point-like, roughly speaking.

A wave represents an ensemble of measurements, not one. One measurement (a point or a spot) is a bit of information about the wave. So in a single experiment it is not a "wave collapse" that happens but information obtaining about the wave. With sufficient statistics you can get an idea about the wave, interference, etc. You can look at this as at necessity to have many bits of information since the studied phenomena (wave behavior) is not elementary (is not reduced to one point while repeating observation).

No, not only but mostly because renormalizations (subtractions) is not a mathematical way of solving equations. It is cheating. I agree formally with your sentence but there are not only aesthetic but physical and mathematical problems. You say it works but nobody can give a picture of a real electron, i.e., of a charge permanently coupled to the EMF variables. Everybody imagines a point-like charge. You see, a working theory cannot say what we deal with in the end. Correct me if I am wrong. So my position is to develop a better theory. I have some ideas and technical implementations but I have not found any possibility to discuss it with researchers. Researchers are happy with the renormalizations and clearly say that they do not need anything else.

It is a sum of the most divergent diagrams. I do not have this book but I think that r0 is not the electron radius. It is 1/Λ where Λ is the cut-off. When Λ tends to infinity, r0 tends to zero and the electron charge gets completely screened at distant r (if it is finite at r0=0). This is a "Moscow zero" - a result obtained by Pomeranchuk and Landau. Although somewhat off-topic, you can find a similar difficulty described in my article "Atom as a 'Dressed' Nucleus" (arXiv, CEJP).

OK so far. I do not like eating a "pudding" made entirely of non eatable stuff (dust, for example). I prefer a pudding made of eatable stuff and in physics we must create and deal with physical entities. Bare particles are not those. Agree, a theory that needs immediately repairing (UV and IR problems) is initially wrong. Wrong, we see the experimental data via our model patterns. Can you outline how you see a real (dressed) electron, please?

Yes, it is. You can find out about it on Wikipedia, I guess. There is a plenty of experiments with different antiparticles and their beams. Soon one will be able to deal with anti-Hydrogen which is especially interesting in many respects. A pale color is not good for emphasizing, in my eyes. Use better bold, italic, and underlined texts of a black color for this purpose.