I am currently in the field helping with some basic avoidance and mitigation work. The way that the process goes is as follows:
Once we have identified archaeological sites, we notify our client and the relevant government agencies (those that issue permits or money or both). The client then generally has one of three options*: determine whether or not the site is avoidable for a state or federal register of historic places, and therefore requires mitigation or avoidance; or just avoid the site to begin with.
I have written about determining site eligibility before. So I'm going to talk about avoidance here.
Avoidance seems pretty straightforward, and sometimes it is. Sometimes it's as simple as "Hey, if you move the road 10 feet to the west over here, then you avoid a site!"
Sometimes it's not. First off, you have to know precisely where the site is in order to avoid it. This is not always simple. For example - if a site is largely buried, then what you see on the surface is just the tip of the iceberg, and determining where the site is, and therefore what must be avoided, requires digging many, many holes and simply seeing if there is anything under the surface. And how deep you have to dig (and how difficult it is to do so) depends on the erosional history of the landscape. For example, I once had to dig a hole that was 50 cm (about a foot and a half) around, and I had to go down 2.5 meters (about eight feet). In the end, most of the upper half of my body was in a hole, while I used two shovels like a pair of claws to pull rocks and soil out of the bottom of the hole. And, despite layers of sterile deposits, I found the buried site at the bottom of the hole.
In addition, you have the question of what the government agencies with which you are working consider to be a single or multiple sites. For example - I worked with one agency archaeologist who wanted a group of five different sites listed as one site, as none was more than 200 feet away from another. A few years later, same agency but with a different archaeologist, wanted all sites that were more than 150 feet apart separated into different sites. So, the boundaries of the site changed not because the site itself changed, but because of agencies direction. When they were gathered together, it might have been permissible to construct where there were no archaeological materials, but it might also not have been, depending on the views of the agency archaeologist concerning the relevant laws.
Then, once you know where the site is, you have to figure out how to avoid it. again, the most straightforward answer is to simply move what is being built, and this is usually what happens. But not always. If a site is to be covered with something such as a temporary access road, or a feature that doesn't require much ground disturbance, it might be acceptable to simply cap the site with gravel. The gravel takes the punishment, and can be dug through if it becomes necessary to access the site later.
A more controversial measure is to cap a site underneath a proposed building. For example, in the town in which I reside, there is at least one major building that has been knowingly constructed on top of a prehistoric archaeological site. The project engineers avoided damage to the site by building up several layers of earthen pads, and placing the building's foundations in these pads, thus avoiding damage to the site itself.
On the one hand, this does protect the site - it is not damaged, and the presence of the building is likely to prevent problems such as site looting.
On the other hand, the legal rationale usually given for protecting sites is that they may yield information important to the study of human history, and are therefore eligible for the National Register of Historic Places under Criterion D: data potential. And, well, if there is a building on top of the site, the odds of successfully studying the site are something close to zero.
One further problem comes in when you consider that a site might be considered eligible for the National Register of Historic Places because of a particular feature or attribute. If this is the case, then there are two schools of thought: 1) the site is eligible, and therefore must be protected in it's entirety; 2) the site is eligible because of a specific aspect of the site, and as long as that aspect is not changed, damage to other parts of the site is acceptable.
I tend to fall into category 1, as do most archaeologists. However, I have come across agency archaeologists who go for category 2, and have been told that it is acceptable to sacrifice certain portions of eligible sites because this will not impact their eligibility in legal terms. There are usually good reasons for this - it allows the protection of some portion of the site, when it would otherwise have been even more heavily damaged - but it always seemed to me to be a bit like using the letter of the law to defeat the spirit of it.
Anyway, I don't know if this is of any interest to anyone outside of my field, but it's a little bit o' the archaeological world that most people aren't familiar with.
*Please note - there are many exceptions