When it is fitted to supernova observations, our standard picture of the cosmos tells us that the current boundary of the universe which can affect us is at z = 1.73. As of today, no event that occurs beyond that limit can ever be known to us or have any causal effect. This is primarily based on empirical observation, not on theory. Whatever the underlying reason may be, we have observed evidence of accelerated expansion. From which this causal limit follows. The question is how to assimilate this into our theoretical model. What has been causing this accelerated expansion? How should we write it into the equations so that they will predict what we have seen? If there are several ways to re-write the equations to be consistent with observation, then which way is simplest and involves the least "made-up" stuff? Easson Frampton Smoot have proposed to include a BOUNDARY TERM in the equation defining the model that would represent whatever is right on this boundary. EFS have now shown that their proposal is approximately consistent with the observed amount of acceleration. So it is a candidate: one of the possible ways of adapting the standard cosmo equations. They add a boundary term, but then they do not need an exotic "dark energy", or an ad hoc "cosmological constant" whose small but positive size has puzzled some people because it looks like fine tuning. However a natural and fairly frequent reaction to the EFS proposal is to say "Isn't this circular reasoning?" So that is what this thread is intended for. Is it circular reasoning? In what sense? And if so, is that bad?