**3. Conclusion**

The history of estimating oil reserves is a history of long lasting misunderstandings. Although American geologists combined volumetric and statistical methods specifically tailored to the realities of the petroleum industry, the final figures from the first oil survey, released in 1909, came to be interpreted by analogy with the forest conservation practices and policies. The discovery of 15 billion barrels of oil left in reservoirs was regarded as a sort of opening shot in a race against the clock of depletion. "Reserves" were understood as a stock; a finite stock that had to be economized, held back and set aside for future uses or contingencies. By adopting terminologies with familiar nontechnical meanings furthermore colored by the moving debate on presidential powers and federal forest "reserves", geologists ascribed the meaning of the concept to an observable fixed asset. Furthermore, given there was, after all, not so much of it left underground, they conveyed the idea that America was approaching its resource supply potential. What ensued was a sort of pathological split between the "scarcity" and the "overflowing" stances, a split sturdily entrenched in discourses, social networks, newspapers, journals and institutions.

To counteract looming claims over the need for regulation, the American Petroleum Association set up its own survey, based on preferential access to oilfields and business records, thereby building a spotless reputation for data gathering. Its aim was to replace geological uncertainties by a narrow but accurate appraisal of oil reserves. Between 1925 and 1935, all open possibilities were locked-into the concept of "proven reserves", grounded on technical and economic feasibility. Such closure ran against the grain of current technological improvements as it excluded probabilistic methods of oil finding by the geophysical sciences and neglected the novel enhanced oil-recovery practices. This means that, in the end, economic-political factors superseded the technical and scientific factors. Comparatively, the Soviet system of oil reserve classification attempted to encapsulate technical feasibility and geological certainty in each category. Indeed, its major originality lay in carving up very analytical classifications with data imbued in meta-data. The fundamental asymmetry between decentralized exploration and central allocation of resources was expected to be overcome by this means.

Overall, the conservation debate continues today with the dispute over whether or not oil production is approaching its maximum level and going to enter into decline soon afterwards, following the downward slope of a logistics curve. Dubbed the "peak oil problem" or "Hubbert's peak" (for an overview see Deffeyes, 2006), its basic assumptions reverse the terms of the debate: whilst in the 1920s it was thought that the oil already found set a ceiling on the likelihood of any new discoveries, the peak oil theory argues the ability to find oil is dominated by the fraction of oil that remains undiscovered.
