If everything in the universe is expanding, how could we detect it since our measuring rods would also be expanding?
You can measure by other means than rulers, for example the force of gravity between two objects would be different, as would their densities. But only the very largest scale objects are expanding, on sizes beyond those of clusters of galaxies; things in our laboratories, our solar system, our galaxy do not partake of the ``universal'' expansion.
Why do some things expand and some things not?
The expansion of the universe is predicted by the equations of gravitation in Einstein's general theory of relativity. Any theory is a model, however, not a completely accurate depiction of reality. To solve the equations simplifying assumptions must be made. For our universe a high degree of symmetry is assumed; this is called the Friedmann-Robertson-Walker model and says that matter is everywhere exactly the same - smooth. In this model there is no laboratory or hydrogen atom or solar system. The expansion is only predicted within this model, which observations show corresponds on average to our universe on very large scales, so we believe it is a valid prediction for the large scale behavior of our universe. If one improves the model, however, by including the presence of a clumpy distribution of matter on smaller scales, e.g. here a galaxy but there none, then the same equations predict that areas where matter is concentrated will not expand.
The universe is expanding because of the Big Bang explosion, right?
Not exactly, depending on what you mean by explosion. We tend to use the word explosion to mean a local event, occurring at a particular point, whose pressure is much greater than that of the surrounding and so causes material to expand rapidly away from the center. This picture fails in two ways for the Big Bang. First, there is no center point; the expansion happens everywhere at once. Secondly pressure doesn't cause the expansion because there is no lower pressure region surrounding the universe; there is nothing to expand into. In fact, in general relativity pressure is a form of energy that effectively acts as mass and the gravitational attraction on this extra mass retards the expansion. One can think of the overall curvature of spacetime as causing the expansion.
Why do we see only redshifts of distant galaxies, not blueshifts?
The shifting of light to lower energies or longer wavelengths (toward the red if we talk about visible light) from distant galaxies is caused by their recession from us under the universal expansion. There are other ways to affect light energies but they are less strong than the expansion. In addition objects moving not perfectly in step with the expansion flow will show Doppler or ``peculiar'' velocity shifts. In fact about 10 galaxies do show blueshifts since their expansion redshifts are smaller than their Doppler shifts. This only happens because they are so nearby that the expansion is weak; they only constitute 0.01% of all cases.
How can we possibly know anything about what happened in the universe billions of years ago?
Basically the same way archaeologists learn about the early Earth - by fossil relics. As discussed in the main chapters there are several relics we rely on. Prominent among these are the light appearing as the cosmic microwave background and the atomic nuclei we detect on Earth, in stars, and between stars. These are concrete evidence of the conditions in the early universe. As well we depend on understanding of astrophysics theory to tell us given the conditions we observe today (and in the past for remember looking at distant galaxies is looking at them as they were long ago) how they must have been early on to have evolved to the present state.
Surely there must be limits to the kind of theoretical extrapolation you talk about?
Very very true. That is why astrophysicists are always hungry for more observations to test, extend, or correct their theories. Theories always must be built on a hard foundation of evidence to be useful in describing reality. Recall there are two components to figuring out the way the universe was long ago - knowing how it is now and how it evolves over time. Imagine scientists examining a glass of water. They see that as its temperature lowers the water's density and other physical properties change slightly, but if they only had ever observed it in the range 20 down to 10 degrees celsius their prediction of its behavior at -10 degrees would be wildly off because they would probably not have guessed at the freezing transition to ice - their theory, based on a limited range of observations, was incomplete and so their extrapolation became invalid when they tried to extend it too far.
What then are the limits of extrapolation for cosmology?
Just as we talked about temperature in the example above, we can talk equivalently about energy or density in cosmology. We have some direct experience of energies at particle accelerators up to about one trillion electron volts, roughly analogous to a temperature of 1016 degrees. We also believe we have a reasonable understanding from laboratory experiments of the behavior of matter at densities up to that occuring in atomic nuclei, about 1014 grams per cubic centimeter. The main extrapolations we make to the early universe that provide support for the Big Bang cosmological model concern the cosmic microwave background and its properties when its temperature was 3000 degrees, and primordial nucleosynthesis which occurred when the temperature was about 1010 degrees and the matter density about 10 grams per cubic centimeter, all well within our confidence range. This is not to say we know everything about what happened back to those epochs, but that we have good reason to believe the broad picture is correct. If we followed our extrapolations only back as far the limits of direct experimental verification mentioned above, this still covers the universe back to the time it was one billionth of a second old. The uncertainties that cosmologists talk about due to quantum effects lie far past this, when the age was a mere 10-43 seconds.
What does it mean to say space is flat?
Physicists distinguish carefully between talking about flat space and flat spacetime; they mean two very different things. Spacetime was introduced by Einstein and is the melding together of space and time into four dimensions. Only under special circumstances can you talk uniquely about time as separate from space. In flat spacetime Einstein's general theory of relativity reduced down to his much simpler special theory; outside of astrophysics this is the theory physicists use when dealing with motions of particles. When the system you are examining, in this case the universe, possesses a lot of symmetry it is sometimes possible to split off the concept of time separately and discuss the curvature of space. Curvature is a geometric concept useful for understanding why distances and forces don't seem to work according to the old Newtonian laws. It turns out to be a much more effective concept than postulating all sorts of new forces. The curvature of space is determined by how much matter and energy are in that region. If there is too little, the space is said to curve negatively, too much gives it positive curvature. The critical amount is determined by the expansion rate and if the total density matches this value then the forces again resemble the ones of Newton using the everyday Euclidean geometry. Why should the critical density depend on the expansion rate? Because the spacetime curvature is given by that and the gravitational attraction of the matter needs to overcome part of that to make space flat.
What is the relationship between the observable universe and the entire universe?
Because light travels at a finite speed and we observe distant objects by means of the light they emit, we can only see objects whose emitted light had time to travel to us since the beginning of the universe. Imagine a postal service that just started up on January 1 that delivers letters over a distance of 100 miles each day. If your friends around the world starting mailing you news as soon as the post office opened, by January 10 your worldview would be everything within 1000 miles. The whole world, however, is much larger. It is the same thing with light in the universe. We can only see out to about 10 billion lightyears distant, beyond this might be a whole part of the universe from which we can get no news. As time goes on we will be able to see from a larger and larger part of the universe, though of course the news will be over 10 billion years old, just like the letters you receive do not tell you how your friends are now, only how they were when they wrote the letters. If the universe is a finite size (what cosmologists call a closed universe), like the Earth is, then eventually we would be able to see the whole thing, if it lasts long enough for light from the farthest object to reach us.
What is meant by the edge of the universe where velocities reach the speed of light?
There are a number of concepts put together in this question. By edge of the universe cosmologists mean the edge of the observable universe or the horizon. Just like on Earth the fact that you cannot see past the horizon doesn't mean there's nothing past there! (See the next question for more details.) This edge is equivalent to the distance light can travel in a time equal to the age of the universe. In an expanding universe objects separate from other nearby objects at a rate given by the Hubble law: v=H0 r where v is the recession velocity, H0 a constant called the Hubble constant, and r is the distance between the objects. If we ask at what distance R the velocity approaches the speed of light c we can solve for R=c/H0. This is known as the Hubble distance. Now the age of the universe is also given by 1/H0 so we see that recession velocities reach the speed of light when at the distance that we had just defined to be that to the edge of the universe. It all seems very neat but in fact there are two complications. We said that the Hubble law is only good for nearby objects, ones much closer than the edge of the universe, because further out the curvature of space begins to upset our notions of distances. Also, the slowing of the expansion changes slightly our age formula above. Thus in general we cannot say that recession at the speed of light is achieved at the edge of the universe. It turns out, however, in the specific case of flat spacetime (called a Milne universe), that these complication disappear and the statement does hold exactly.
How could an infinite universe have been, say, the size of a golf ball?
This is another good example of the case where scientists use words in specific ways, leading astray those who read those words with their general meaning. The ``infinite'' in the question really refers to the geometry or topology of the universe as a whole, while the word ``size'' refers to the extent of the horizon or visible universe (had there been an observer at that time). There are three basic geometries our universe could have been born with - closed (spherical), flat (planar), and open (pseudospherical). Even at the time the observable universe was just a point, the entire universe still possessed one of these topologies. Only in the closed case is the full universe finite. Imagine holding a flashlight close above a large piece of paper. As you raise the flashlight steadily over time, the illuminated circle grows but this does not affect size of the actual paper.
The inflation scenario predicts the universe expands faster than the speed of light. Doesn't this violate the laws of physics?
There are many different inflation scenarios but in fact the unifying definition of inflation is that it involves an era of superluminal expansion. The laws of causality, i.e. cause and effect, say that no form of energy or matter can travel faster than the speed of light. Nonphysical things, basically mathematical concepts or figments, are not so restricted. What is said to expand in inflation is the particle horizon, which has no physical form or structure but is instead merely a useful concept like the equator on Earth.
Why does our universe seem to contain only matter and not antimatter? Could other galaxies be made of antimatter? Could a universe of only antimatter exist?
Actually our universe does contain antimatter also, though in very small amounts and generally only near the site of very violent, energetic sources. When matter and antimatter come in contact they annihilate each other, producing very characteristic radiation. Because galaxies are not totally isolated from each other - the space between them is not totally empty just much less dense than the galaxies themselves, also sometimes galaxies collide with each other - we would expect to see this radiation if there were antimatter regions of the universe. Another problem is how the matter and antimatter would have so neatly segregated themselves to begin with. One thing to note is that matter and antimatter are not exact twins of each other, there are very slight asymmetries in their properties. One theory of the generation of matter in the early universe says that matter and antimatter were created mixed together but because of the slight difference they did not fully annihilate each other but left a small excess of matter. Predicting whether this excess matches certain observational clues from the early universe is a exciting frontier of current cosmological research. Until we understand the generation mechanism better it does seem possible that in theory a wholly antimatter universe could exist, though again because of the slight asymmetry it would not be an exact twin of our own universe.
Why is the night sky dark?
This is a classic question, often known as Olbers' paradox. The brightness we see is the brightness of whatever object emitted the light entering our eye along that line of sight. To get a dark patch of sky one needs there to be either no objects in that particular direction, or have only dim ones. Stars are very bright so if there was a star along that ray we would expect a bright patch unless there was some way to make it dimmer. Distance alone does not dim an object, it just spreads the energy out more. However, the redshift due to expansion does reduce the brightness somewhat. Thus there are two possible solutions to the puzzle - the night sky is dark either because on average there are no stars along the line of sight or because the expansion dims their brightness. In our universe it turns out that both effects enter, in different areas. For objects like stars, there are too few within our horizon to give good odds for a particular ray to hit a star. Since the horizon depends on the age of the universe it is often said that the solution is that our universe is too young. For a spread out background glow like the cosmic background radiation, expansion is the main effect, dimming the brightness over time.
(c) Eric V. Linder 1993