In John Lewis Gaddis' The Landscape of History: How Historians Map the Past, the author uses the metaphor of mapping a landscape to illustrate how historians try to map the past. For example, if a historian measures the distance between two continents in miles, he or she will arrive at one figure. If her or she measures the distance in micro-millimeters, it is likely all the nooks and cubbyholes will be discovered and the figure will be much higher. Both provide "maps" of the distance, but each provides a different level of depth in the portrait of the space. Portrait is an apt word, since Gaddis maintains historians are like artists or scientists in their attempt to recreate the past in order to provide broader understanding in the present. Sociological approaches to history are too rigid, dependent on variables and oversimplify the past. Scientific approaches are more akin to the historian's methods, especially a degree of uncertainty and unpredictability. In essence, Gaddis argues that the best historians can do to represent past reality is to ensure it is portrayed with an "acceptable" degree of error, much like scientific discoveries. A conclusion will address how historical method allows us to know the past as much as we can know it.
Sociological approaches and formerly even scientific approaches are too mechanistic in Gaddis' view to ever contain or portray with any accuracy the chaos and complexity that are history. The best that can be done under any circumstances to portray history is a kind of "mapping," a process that is similar to sciences like cosmology or paleontology as well as history in that only a representation can be presented. In many ways, the mapping of history is simultaneously a linear as well as chaotic process and only approximates truth (with varying degrees of error). Gaddis maintains this is not a new discovery but one Poincare defined that also applies to history: "Poincare...