The first step in our study of aesthetic standards in source code will identify the aesthetic ideals ascribed by programmers to the source code they write and read; that is, the syntactic qualifiers and semantic fields that they refer to when discussing program texts. To that end, we first start by clarifying whom we refer to by the term programmers , revealing a multiplicity of practices and purposes, from massively-distributed codebases to ad hoc , one-line solutions, cryptic puzzles and printed code.
We then turn to the kinds of beauty that these programmers aspire to. After expliciting our methodology of discourse analysis, we engage in a review of the various kinds of publications that make up programmers' discourses, in which they qualify their practice. Out of these, we identify a cluster of adjectives and comparisons which will provide an empirical basis for considering the desirable and undesirable aesthetic properties of source code.
We then move to a description of which aesthetic fields are being referenced by programmers on a broader level, and consider how multiple kinds of beauties, from literary, to scientific and architectural conceptions of beauty can overlap and be referred to in the same medium. Such an overlap will reveal the importance of function, craft and knowledge in the disposition and representation of code. Our conclusion focuses on how understanding plays a central role in an aesthetic approach to source code, and results from the specificity of code as a cognitive material.