The Failures of Mathematical Anti-Evolutionism
vertex set and E(G) the edge family
Download 0.99 Mb. Pdf ko'rish
|
The Failures of Mathematical Anti-Evolutionism (Jason Rosenhouse) (z-lib.org)
V(G)
the vertex set and E(G) the edge family of G. (Wilson 1996, 9) That sure seems a lot more complicated than what we discussed in the last section! Why did Wilson not just say, “A graph is a diagram in which you have a bunch of dots connected by a bunch of lines”? After all, that is pretty much how we defined things in Section 3.1. The answer is that a definition solely in terms of dots and lines leaves too many unanswered questions. How many dots do you mean? Do we want finitely many dots, or are we allowing infinitely many? Are the lines allowed to start and end at the same dot? Do we want to think of the lines as arrows, so that there is a preferred direction in traveling from one dot to the next? Do we want to allow multiple lines to connect the same set of dots? In the problem of the Königsberg bridges, we did allow multiple lines because in some cases the same two land masses were connected by more than one bridge. On the 3.2 you need both rigor and intuition 67 other hand, if the dots represent cities and the lines represent the existence of a direct airplane flight between them, then maybe we only want to allow a single line. Come to think of it, what do we even mean by dots and lines? In the brief list of applications we mentioned in Section 3.1, what we cared about was the relationship of what was paired with what. As we have discussed, the picture with the dots and the lines helped us to visualize and understand the graph. But just as the map is not the territory, the diagram is not the graph. If we erase the drawing from the page, the abstract relationships will still exist. These are the reasons we needed a precise definition of the sort that Wilson provides. Notice that he anticipates and answers all of the questions I asked in the previous two paragraphs. He stipulates that we only want finitely many vertices. He emphasizes that the edges are unordered pairs, meaning that we are imagining lines and not arrows between the vertices. The phrase “not necessarily distinct” implies that it is fine for a line to start and end in the same place. (Such a line is called a “loop” in the graph.) And so on. Our interest here is not really in the technical minutiae of how a graph is defined. We will not see graphs again after this chapter. However, I have belabored these points as an illustration of the parallel tracks of mathematical reasoning. Our examples have shown us two different ways of discussing mathematical concepts. On the one hand we have an intuitive understanding of our objects, of the sort presented in Section 3.1. I think of this as track one. On the other hand we have the rigorous, precise definitions with which mathematicians undertake their professional work. We shall call this track two. As we proceed, we will often make distinctions between track one and track two mathematics. So let us be explicit about what we mean: • Track One: Our intuitive understanding of what is really going on. • Track Two: The logically rigorous and precise technical definitions and theorems commonly found in high-level mathematics books. 68 3 parallel tracks Both tracks are critically important. If you try to get by solely with an intuitive, track one understanding, then it is too easy to make logical oversights when manipulating your abstract objects. Ultimately, everything you know about the objects is contained in their track two definitions, and that is why the utmost precision is necessary in formulating them. But the rigorous, track two definitions are sometimes so abstract that it is difficult to think clearly about the underlying objects. That is why you need track one as well. For example, if you only think of a graph as an abstract set of vertices combined with an equally abstract set of edges, then you would probably never notice that problems about Eulerian walks are intimately related to the parity of the vertices. It is hard to notice that until you think in terms of dots and lines and literally imagine walking around the diagram using the edges as bridges between vertices. But if you only think in terms of dots and lines, then you will not be able to write down a convincing proof that Eulerian walks exist precisely when no more than two vertices have odd degree. That is why both tracks are necessary. Let us push this just a little farther. We shall restrict our attention to so-called connected graphs, meaning that given any two vertices, it is always possible to walk from one to the other. We do not want a situation where one part of the graph is entirely cut off from some other part of the graph. It will also be convenient to consider Eulerian walks that start and end in the same place since that way we do not have to include the proviso that the first and last vertex are permitted to have an odd degree. To indicate that the starting and ending points are the same, we shall speak of an “Eulerian cycle” instead of an Eulerian walk. Now, based on our experience with the Königsberg bridges, we have a good plausibility argument that Eulerian cycles through a graph exist precisely when all of the vertices have an even degree. How do we turn our plausibility argument into a proper proof? First, we will need precise definitions of the terms involved. Continuing with Wilson’s textbook, we shortly come to this: 3.2 you need both rigor and intuition 69 Given a graph G, a walk in G is a finite sequence of edges of the form v 0 v 1 ,v 1 v 2 , . . . ,v m −1 v m , also denoted by v 0 → v 1 → v 2 → · · · → v m , in which any two consecutive edges are adjacent or identical. … A walk in which all the edges are distinct is a trail. If, in addition, the vertices v 0 ,v 1 , . . . ,v m are distinct (except, possibly, v 0 = v m ), then the trail is a path. A path or trail is closed if v 0 = v m , and a closed path containing at least one edge is a cycle. (Wilson 1996, 26–27) If you have no prior familiarity with the concept of a walk in a graph, it would be hard to comprehend the concept based solely on this definition. But if you already understand that you are just moving from dot to dot along the lines of a diagram, it becomes much easier to parse what this definition is saying. After later stipulating that an Eulerian cycle is a cycle that con- tains every edge exactly once, and stipulating that a graph possessing an Eulerian cycle is said simply to be Eulerian, we come to a properly stated theorem: Download 0.99 Mb. Do'stlaringiz bilan baham: |
Ma'lumotlar bazasi mualliflik huquqi bilan himoyalangan ©fayllar.org 2024
ma'muriyatiga murojaat qiling
ma'muriyatiga murojaat qiling