If we think about the way people's lives progress, we all share a common human experience: birth and death. Our direct ancestors must have survived long enough to have children or we wouldn't be here. A standard pedigree chart shows four or five generations, with specialized charts going back to ten generations or more. This is the framework of the structure upon which you build your genealogy. There is a difference between claiming a brick wall for one of you direct line ancestors (grandparents, great-grandparents, great-great-grandparents etc.) and claiming a brick wall on a collateral line, such as the sibling of a grandparent, the spouse of a sibling or whatever.
Collateral relatives are those to whom you are related by blood, but are not your direct line ancestors. Therefore, your ancestors are your parents, grandparents, great-grandparents, etc., and your collateral relatives are your cousins, nieces, nephews, aunts, uncles, siblings, etc. Now, you are usually not related by blood to your spouse (although you may be if you married a cousin). Likewise, you are not blood relations to the spouses of your collateral relatives. For example, I am not related to my great-uncle's wife the same way I am related to my uncle. My uncle and I share a common parent (grand-parent) but his wife has an entirely different set of blood line relatives who may or may not be my relatives.
Now, at this point, I need to discuss pedigree collapse. This is an inevitable and unavoidable, basic fact of genealogy. As we go back in the past our pedigree takes the shape of a snake that has swallowed some large animal. If you think about it, your number of ancestors doubles in each generation: 2, 4, 8, 16, 32, 64, 128, 256, 512, 1024, etc. If you were to go back thirty generations you would, theoretically, have 2 to the 30th power direct line ancestors or well over the entire population of the world at that time. So what happens? We all end up marrying cousins. Eventually, the pedigree line collapses and the number of real direct line ancestors begins to decrease rather than increase. For example, the offspring of two first cousins have at most only six great-grandparents instead of the normal eight.
One problem with using a standard genealogical database program is that these programs do not adequately reflect pedigree collapse. For example, my parents were second cousins. If I go to FamilySearch.org's Family Tree and show my pedigree in a fan chart, it will show four generations of my direct lines. What it does not show is that my mother's grandfather and my father's great-grandmother were brother and sister. So, from a practical standpoint, if I am working off of a standard fan chart or pedigree chart, I may not even realize that the lines cross.
Now, as you build your genealogy on a direct line, you take advantage the of the research that you do at each generation. For example, as I investigate the lives of my parents, I find information about their parents and so forth. If you try to skip a generation, you may get lost. First, you are ignoring the information you could have obtained from the descendants of the direct line ancestor and second, you are making an assumption that the person you are investigating or researching is the correct person. As an example, let's say I have inherited a pedigree chart from someone in my family. I notice that going back a couple of generations, there is a blank space for my paternal great-great-grandfather, I immediately decide that I want to find him and start doing research.
I cannot tell you how many times this occurs. Unfortunately, the person doing the research has no idea who the great-grandfather is and does not even know if the right person is recorded. This is like trying to build a bridge from the middle out of the two shores. There is no anchor or foundation for the research. Now, what you do when you start researching the spouse of a collateral relative is exactly the same thing. You are building a bridge in the air. You have done no research at all on the children of that spouse and you have no idea if the spouse is correctly identified. So you don't start with the missing spouse, you start with the children or grandchildren.
You can't start doing effective research without the foundational research that provides a base for the investigation. You may find the person or you may not. But if you have failed to do foundational research by examining in detail the children of this collateral relative and his or her spouse, then you are building a bridge in the air. Why do you think I find so many of the children of my ancestors assigned to the wrong wife of my collateral relatives? It is because the researchers start doing the research with the spouse and ignore the research on the children.
Come forward. This is one of the keys to finding an ancestor. It may seem like a waste of time to do the research on each child and each grandchild, but this is exactly what needs to be done. You have to know who you are trying to research. The difference between claiming a brick wall for a direct line ancestor and a collateral ancestor is simple; the amount of research you have already done on the person's descendants. Usually, researchers have done little or no research on the descendants of their collateral ancestors.