Dark matter and dark energy are typical examples of what is known in the philosophy of science as “theoretical entity”: elements of theories, whose existence is hypothesized and with assumed properties and behaviours that can be used for explanation of the observed phenomena. What is regarded as an observable phenomenon, and what as a hypothetically assumed theoretical entity, depends on the state of theory and the empirical degree of assurance. Most of the phenomena of modern physics or cosmological theories are themselves – at least in an everyday sense – not observable, but by the diversity of empirical evidence and experimental investigation of the existence of these phenomena is relatively certain.
For example, you can consider the cosmic background radiation and the expansion of the universe as highly theoretical constructions, theory elements that make it possible to explain results of radio telescope measurements and spectral analysis. The multiplicity and diversity of these empirical findings, however, lead in the course of the research to the suggestion that such descriptions can be regarded in fact as observed phenomena. Although we can never be completely sure about its actual existence, the certainty that they exist and that they are covered by our descriptions, increases with each new experimental result which ifits the picture that we have made of these things.
For the “dark energy” it is still a long way from being a theoretical entity to becoming an observable phenomenon – but maybe with the work that Christian Marinoni and Adeline Buzzi have published in late November in nature  on this path is once again done a little step.
The idea of dark energy has been introduced in order to explain the phenomenon that the expansion of the universe accelerates at present. Of course we can not observe this acceleration in the same sense as the acceleration of a car. Nevertheless, we consider the acceleration of the expansion of the universe today as an observable phenomenon, there are a number of observations that fit well with this picture.
The “dark energy” shall explain this acceleration – namely because the gravity would actually slow down the expansion of the universe, there must be something that works against gravity – and this something was called “dark energy”.
As long as the dark energy’s only benefit is to explain the acceleration of the expansion of the universe, it is a weak concept. If fact, one would have only another name for the phenomenon. We must therefore find out more about this hypothetical entity, we have to find other traces of its existence, traces that are regardless of the expansion acceleration.
One such idea is introduced by Marinoni and Buzzi. They wanted to find a way to test the cosmological model, which includes the dark energy, by developing a measurement method for the geometry of the universe. The idea is to examine pairs of galaxies, i.e. galaxies that orbit around their common centre of gravity. The expansion-related red shift must then be different depending on the current orientation of the axis of the system for the two galaxies – and the assumed correct “content of the Universe” – which includes the right amount of dark energy. Only when the contents of the universe are assumed correctly, the calculations will agree with actual measurements.
The accuracy of the results is certainly not yet sufficient to take them as evidence for the assumptions about the dark energy – but Marinoni and Buzzi have shown that their way is basically functioning. The accuracy will increase over the next few years – and thus, a new method of “observation” of the dark energy can be established. And with each new established procedure increases the certainty that the dark energy really exists as such, as the researchers think of – it will be developed step by step from a theoretical entity to observable phenomenon.
 Marinoni C, & Buzzi A (2010). A geometric measure of dark energy with pairs of galaxies. Nature, 468 (7323), 539-41 PMID: 21107424