Conversation analysis (CA) is now what Kuhn once called a normal science. It has a discernible body of concepts, methods, and recognizable objects of analysis. More importantly, its considerable archive of accumulated findings has a very high degree of redundancy–in the positive sense that researchers have continually replicated the findings of their colleagues. It ought, then, in every respect, to be the envy of the social sciences generally and not easily dismissed as an abstruse and recondite branch of language studies or "discourse analysis" concerned with merely trivial fragments of recorded talk as such. To help CA be respecified in this way, the present paper unearths a relatively unknown part of its history in the partial book manuscript of Harvey Sacks's Aspects of the Sequential Organization of Conversation, circulated in mimeo to colleagues in 1970. There we find innumerable appeals to the now well-recognized fact that the practical actions of everyday talk have significant, culture-wide, formal structures that are the proper objects of the discipline. Or, in Sacks's own terms, seemingly insignificant mundane events are generated by a powerful (if not always intuitively accessible) social "machinery." The purpose of this paper is, then, to extract a sample of these aspects of Sacks's Aspects in order to illustrate, sketch, or give a preliminary sense of just that point.