Slow roll inflation -layperson question

[a dull boy]

Dear PF,

I'm trying to understand slow roll inflation from a layperson view, and I think my current understanding is mistaken. If a bubble of matter is trapped in a local minimum PE state, and begins to roll down a hill to a lower PE state, it strikes me that bubble will expand in size the same way a bubble of air trapped in water expands when it reaches the surface. But I don't think this is at all how cosmologists explain inflation. Could someone help?

Thanks, A dull boy

 

[bapowell]

What exactly do you mean by "when it reaches the surface"?

The early conception of inflation indeed was understood as a bubble nucleation event: the inflationary phase transition was "first-order" so that the field jumped spontaneously from the state of higher PE to vacuum, creating bubbles of vacuum in the inflating region. The bubbles would then expand outwards at near the speed of light, converting higher PE regions to vacuum. Bubbles then coalesced to complete the inflationary event in our observable universe. As you might know, there are serious problems with this sequence of events, namely that the resulting universe would be strongly inhomogeneous on account of the bubble collisions.

To address this problem, "new" inflation was quickly proposed by Linde and others. Here, the transition is not first-order, but second-order, or at least only very slightly first-order. Instead of jumping directly to the vacuum, the field either slowly rolls off the local maximum towards the vacuum (in the case of second order), or tunnels through the barrier and then slowly rolls to the vacuum (in the case of lightly first order). Here, it doesn't help so much to think in terms of bubble nucleation, although that is very possibly what is happening (the Hawking-Moss instanton describes bubble nucleation appropriate to second-order transitions). The reason is that the whole of the observable universe after inflation fits inside a single bubble; this is because the inflationary phase transition takes so long to complete (it is a "slow" roll after all). This solves the problems of old first-order inflation by avoiding the need for violent bubble collisions.

So, in summary, while the slow roll phase transition does likely occur inside an expanding bubble (where the PE inside the bubble is decreasing more rapidly than that outside), because the whole of the observable universe fits inside the bubble, we need not talk about bubbles or worry about their dynamics. We can instead work within the bubble where the universe is effectively infinite.

 

[a dull boy]

Thank you very much, I understand this much better now.