Representing and Searching the Space of Mathematical Knowledge We discuss search in highly modular theory libraries. We represent such libraries as OMDoc/MMT theory graphs and host them in the MathHub.info system, a portal for active mathematical documents and an archive for flexiformal mathematics. In theory graphs, statements can be inherited (and thus need not be explicitly represented) via morphisms that can include translations. This is good for knowledge management as the number of induced (i.e. not explicitly represented) statements can grow exponentially in the explicitly represented ones. But this is bad for accessibility, since conventional (computer supported) access methods only work on explicitly represented materials. We claim that mathematicians are usually aware of the modular structure of the mathematical domains they understand and induce the inherited knowledge space from the documents they read and the discussions they participate in. In particular they conceive mathematical domains as the induced knowledge space and can access it independently of how it was represented in the sources. We call this access pattern ``math-literate'', and observe that regular search engines for mathematics (including our own MathWebSearch engine) are not, because they cannot make the necessary coersions induced by theory morphisms. In this paper we note that if the representation framework of the modular library provides a systematic naming scheme for induced statements and the library system implements a flattening operation that generates all induced statements, then we can use this for access. We extend the MathWebSearch System with index flattening and integrate it into MathHub.info as a math-literate access system.