This paper examines some soils in Michigan's upper peninsula that do not "fit" with the zonal soils of the region. The typical upland soils in this part of the Pictured Rocks National Lakeshore (PRNL), when not constrained by bedrock or high water tables, are strongly developed Spodosols (Haplorthods). We used the energy model (Runge, 1973) as a conceptual guide to help explain the genesis of nearby soils, whose chemistry and morphology are very different. The energy model presupposes that soils are affected mainly by water available for leaching and organic matter production, as conditioned by parent material. In our study area, where water available for leaching is limited (or even negative), as on steep slopes shallow to bedrock with multiple springs and seeps, and where erosional processes are strong, horizonation in the classical sense does not form. Instead, shallow Histosols (Saprists) form above bedrock, even on 45% slopes. Oxyaquic Haplorthods developed on flat uplands, on the same bedrock, are strongly horizonated, as most of the water that impacts them is available for leaching and pedogenesis. In another example, soils that have formed above calcareous bedrock maintain such high pH values that their faunal assemblages are rich and diverse, and base cycling is strong, leading to thick O + A horizon sequences. These soils (Lithic Udipsamments) resemble Mollisols and exhibit few marks of podzolization. Conversely, nearby Oxyaquic Haplorthods above acid sandstone bedrock have limited faunal assemblages, slow organic matter decomposition, and horizonation indicative of strong podzolization. These examples highlight how conceptual models can be used to guide our understanding of soil genesis and distribution.