I consider the role of the source code of generative literature in the process of meaning making. The significance of code in the cultural meaning of generative works means the source code becomes a key factor to explore in literary studies. I use Critical Code Studies (Marino) which rejects the practice of only analyzing the output of electronic literature and instead proposes to look at code from a humanities perspective as an integral part of coded literature. To specify this emerging field specifically for generative literature, I propose a distinction between three levels on which the code is involved in the meaning-making process of generative literature: the linguistic level, the literary level. and the cultural level. On the linguistic level, I draw from structuralism, using Jakobson's notions of selection and combination as outlined in "Two aspects of language and two types of aphasic disturbances". Generative literature shows the meaning of language explicitly via selection and combination of linguistic units, and adds to this process a literary meaning employing the process of chiasm and overwriting. To do justice to the complexity of the materiality of coded literature on a literary level, I link this to Brillenburg et al's reference to Lyotard's notion of chiasm as excess of meaning and Dworkin's notion of neglected perspectives. Moreover, the source code is positioned as a trope for objectivity, as it does not embody the same cultural biases as one expects from intention-typical research. On a cultural level, I argue that source code is positioned as a trope of objectivity, as the randomness of generation supposes an emptiness of cultural bias.
(author abstract)