this work intends to demonstrate the role of aesthetics in source code in understanding a program. this demonstration will resolve the contradiction that, on one side, understanding is multi-faceted and, on the other side, there are language-independent and domain-independent practices which facilitate understanding.
the significance of the study is the possibility to adopt close-reading practices to a body of texts which has so far remained largely unexplored. it is using both cross-disciplinary frameworks to further qualify a somewhat new and unique kind of text.
this study contributes to a finer understanding of programming as a a practice in-between engineering and literature; it creates this middle-ground not at the higher level of general concepts, but at the practical level of reading and writing. this practical approach ties the notions of understanding that we explore within a broader socio-economical context, and as such requalifies both engineering and literature as, respectively, sociotechnics and socioaesthetics.
there has been no significant study of the aesthetics of source code from a cross-disciplinary perspective. either you have cross-disciplinary (software studies, etc.) and they do not (or very little) (or through the poetics, not the aesthetics) address source code directly, or you have computer science and they only address it in terms of their own particular practice, sometimes reaching out to “literary” or “beautiful”, but without extensive theoretical backing. this thesis proposes to have both enter in a dialogue. a final component of the problem at hand is that there is little intertextual analyses of source code from a literary perspective, so far contained to text-by-text readings rather than more holistic approaches.
the methodology is inductive, and qualitative. by gathering and reading primary sources, i establish the practice-based aesthetic standards to which practitioners submit their work. i propose an inductive approach first because there is no “criticism” in programming, but rather “peer-reviewing”, which means that anyone who writes, reads, and therefore what the writers and the readers say is equally important. second because the cross-disciplinary approach of analyzing those texts inherently forbids the blanket use of one theory mapped onto possibily different practices. rather, establishing a set of practices and then invoking theoretical frameworks to further qualify these practices avoids the trap of overfitting.
(1) i am assuming that aesthetics (including the lack thereof) is an essential part of source code. since it is not possible to separate both, every text becomes a valid means of examination since it would display some dimension of aesthetic properties. of course, some texts might be more useful than others. (2) i am assuming that writing practices are influenced by additional writing published in the field, from the art of computer programming to blog posts and best-practices lists (kinda theory of communicative action).
in terms of limitations, the clear distinction between closed-source and open-source will inherently limit the scope of our body of text. (2) there is also my limited understanding of certain languages and certain hardware. (3) the social component of the study also implies that i might not have access to specific language communities (i.e. hacking and educating in non-english or non-french languages), but that would be mitigated by the fact that most resources are based on english
with this work we propose to demonstrate the thesis that aesthetics play a role in understanding code; the previous research has left many questions unanswered, and the data gathered is still insufficient.