> What are the error bounds on measuring such an object at the distances they exist at? Let's look at two ways of judging the size of a quasar. We have photos

> What are the error bounds on measuring such an object at the distances they exist at? Let's look at two ways of judging the size of a quasar. We have photos of quasars, each at the middle of the galaxy they are at the centre of. We know the distance to the galaxy (and to the quasar) and can measure the size of the galaxy by the arc length it covers. Our errors on these measurements come down to how accurately we can measure the distance to galaxy and its size, but the quasar is very obviously much smaller than the galaxy. The other way is to look at how fast the luminosity of the quasar changes. If an object is observed to change at a certain rate (in the case of quasars, over the course of a few days sometimes) we have a limit on how big the object is. Looking for an upper limit, let's assume that the entire object changes at the same time; some process coordinates the change so that the whole thing gets brighter or dimmer at the same time. The light from each part of the object travels to us and arrives at different times. The parts that were close to us arrive first, the ones farther away last. Even an instantaneous change will be smeared out from the perspective of an observer, the bigger the smear the bigger the object. Importantly, the object (at least the light producing parts of the object, the parts we care about) can't be larger than implied by the smear. If it was bigger, the smear would be bigger. A smaller object can have a bigger smear, by slowing changing over time, but a bigger object can't have a smaller one. Quasars often change luminosity significantly on the order of weeks or days, so we know that they are no bigger than a few light-weeks or light-days across. For reference, our solar system is at least a light-day across, and by some definitions more than a light-year in size (or roughly half the way to the next star). > We make assumptions all the time when we measure things and often we do not question whether what we are measuring is actually correct. We assume. In these cases, we have to use proxies to determine distances and sizes. if any of those proxies are even a little out, our errors can be huge. Who is we in this case? Sure assumptions are made, all the time. When astronomers first started trying to measure distance to galaxies they assumed that all similar looking galaxies were the same size (the 'standard ruler' method). No one thought this was absolutely correct, but it was useful and allowed measurements to be made. Once you have those measurements you can compare with other measuring techniques and see if they agree. Sometimes changes will be made, but the errors are very very rarely 'huge'. As you walk towards a distant mountain, you are better able to see the slopes and craggy outline - it doesn't just disappear. > I am suggesting that the basis assumptions be looked at to see what else may be more applicable. If you do this, valuable insights may be gained and a different approach suggested. This kind of activity is undertaken for simple system like computer models. Why not do this for other such areas as well? Which 'basis assumptions' are you asking to be revisited here? In the article they walked through some of different theories that have been proposed to try and address the mechanics that underlie these objects. > Going back to my original comment, as the model does not accord with the data, maybe we need to be looking at the basis assumption that are being used. But the data does accord with the model! Up until this observation was made, the data very much did agree with what the model predicted. A new observation came up that showed some predictions were incorrect, but the model didn't suddenly become meaningless. Relativity showed that Newtonian physics didn't accurately describe fast moving objects, but Newtonian physics still accurately predicts almost everything else. > Don't be afraid t revisit old ideas or look at new ideas. There may be gems that will open up completely new areas of investigation and discovery. I don't disagree at all, but simply saying we need to reconsider our 'basis assumptions' doesn't do any of that. We have incredibly successful and accurate models, and are excited to find situations in which they don't hold up as that is an opportunity to improve our understanding of the world. But you go and investigate that specific failing, reworking the model to account for the difference. It is a rare and special thing to have to fundamentally change a model based on new observation, and there is no reason to suspect such a change is needed here, yet. Approaching any discrepancy between a model's prediction and an observation with "let's revisit our entire understanding of this subject" is neither useful nor scientific.