You’ve got to love physics humour. To mark Valentine’s day last week, Fermilab’s in-house magazine ran a spoof personal ad that stated: “mature paradigm with firm observational support seeks a fundamental theory in which to be embedded”. It was referring to inflation — a period of exponential expansion thought to have taken place 10–35 s after the big bang, which, although able to account for the large-scale appearance of the universe, lacks a firm theoretical footing.

By chance, that same day three theorists posted a paper on the arXiv preprint server that could help remedy this situation. Viatcheslav Mukhanov and co-workers at Ludwig Maximilians University in Munich, Germany, have proposed a model in which inflation is driven by vector fields as opposed to scalar fields as it is in existing models (arXiv:0802.2068v1). Although their model does not fundamentally explain inflation, vector fields are already known to exist in nature whereas scalar fields are not.

Smooth and flat

Inflation, which was developed in the early 1980s, smoothes out anisotropies that were present immediately after the big bang by causing the universe to expand by a factor of least 1030 in a fraction of a nanosecond. Without inflation, it is difficult for cosmologists to explain why the universe looks roughly the same no matter which direction they look or why the geometry of the universe is essentially flat.

But the underlying theory is somewhat ad hoc. In most models, the rapid expansion of the early universe is driven by “negative pressure” produced when the potential energy of a scalar field drops from one state to a lower one. Quantum fluctuations in this field, with which is associated a fundamental particle with zero spin called the inflaton, would have been blown up to cosmic scales and produced density perturbations that later caused matter to clump together into galaxies.

“Ours is not a better model than scalar-field inflation, but it is also not worse.” Viatcheslav Mukhanov, Ludwig Maximilians University

This picture has recently gained strong support from measurements of the distribution of hot and cold spots in the cosmic microwave background. But physicists have never seen spin-zero particles and therefore have no real clue about the microscopic origins of inflation.

On the other hand, fundamental vector fields — which lead to spin-one particles — are known to exist in nature: the photon and the W and Z bosons, for example. The trouble is that vector fields tend to “pick” a preferred direction in space and therefore ruin inflation’s biggest selling point. Mukhanov and colleagues have now overcome this problem by invoking several vectors which, when averaged, lead to a nearly isotropic universe.

Random orientations

According to Larry Ford of Tufts University in the US, who in the late 1980s came up against the problem of anisotropy himself when attempting to build a vector-inflation model, the new model is analogous to the pressure in a gas. “A gas consists of many molecules, each with a specific velocity, but when the effects of the various molecules are averaged the result is a pressure that is isotropic to a very high degree of accuracy,” he explains.

“Ours is not a better model than scalar-field inflation, but it is also not worse,” says Mukhanov, who was one of the first to calculate the fluctuations in the inflaton field. “In addition, the model allows us to get a certain amount of anisotropy during inflationary expansion which could modify the produced perturbations and thus have observational imprints in the cosmic microwave background.”

Because the model requires many vector fields orientated in random directions to produce the required isotropy of the large-scale universe, its predictions depend heavily on the way the masses of these fields are distributed, which is not a desirable property. The upside is that the presence of fields with extremely small masses naturally accounts for the current phase of cosmic acceleration as well as inflation, although team member Alexey Golovnev is quick to point out that scalar fields can also do this job.