There exists an extensive literature on vision science, on the one hand, and on program comprehension, on the other hand. However, these two domains of research have been so far rather disjoint. Indeed, several cognitive theories have been proposed to explain program comprehension. These theories explain the processes taking place in the software engineers’ minds when they understand programs. They explain how software engineers process available information to perform their tasks but not how software engineers acquire this information. Vision science provides explanations on the processes used by people to acquire visual information from their environment. Joining vision science and program comprehension provides a more comprehensive theoretical framework to explain facts on program comprehension, to predict new facts, and to frame experiments. We join theories in vision science and in program comprehension; the resulting theory is consistent with facts on program comprehension and helps in predicting new facts, in devising experiments, and in putting certain program comprehension concepts in perspective.
In theory, there is no difference between theory and practice. But, in practice, there is. —Jan L. A. van de Snepscheut
TopIntroduction
Joining vision science and program comprehension provides a theoretical framework to explain how software engineers understand programs and, thus, to explain known facts, to predict new facts, and to set up experiments on program comprehension.
In the recent year, the domain of cognitive informatics as emerged to study the internal processing mechanisms of the human brain and their applications in computing. Cognitive informatics is intrinsically multi-disciplinary and unites researchers in several domain of research such as cognitive science, cybernetics, and software engineering. In particular in software engineering, it could bring interesting advances that could explain the mechanisms trough which software engineers understand, write, and debug programs. Therefore, it encompasses the domain of program comprehension.
Program comprehension is the domain of software engineering that seeks to explain how software engineers understand programs (Rugaber, S., 1995), how they obtain a mental representation of a program structure and function (Navarro-Prieto, R., 1998). Facts and laws of program comprehension lie at the heart of almost every software related activities, from development to maintenance, deployment, and use. Many studies on program comprehension have been published in the literature. However, understanding program comprehension requires more than just knowing facts; it requires a theory.
Our research concerns situations in which expert software engineers readily recognise well-known patterns (any kind of patterns) in a program model, while novice software engineers only perceive their constituents. Figure 1 shows a subset of a class diagram. (We cloud class names for the sake of explanation). The question is: What is the structure and function that an expert software engineer recognises instantly in that diagram and that a novice software engineer does not? (The curious reader may take few seconds to study the class diagram.) An expert software engineer could recognise that the classes follow the solution of the Composite design pattern (Gamma, E., Helm, R., Johnson, R., & Vlissides, J., 1994), thus assigning the function of representing part–whole hierarchies, while a novice software engineer only sees a hierarchy of four classes, some operations, and an aggregation relationship. (Compare with the solution of the Composite design pattern in Figure 4). We attempt to explain this difference in comprehension between expert and novice software engineers with a theory.
Figure 1. What is the structure and function that an expert software engineer recognises instantly in this subset of a class diagram of a well-known program and that a novice software engineer does not? How does the recognition happen?
Figure 4. Original class diagram of the Composite design pattern (Gamma, E., 1984)
Existing cognitive theories1, such as Brooks’ (1978), von Mayrhauser’s (1995)Pennington’s (1987), Soloway, E., Pinto, J., Letovsky, S., Littman, D., & Lampert, R. (1998), provide explanations on the short-, long-, and working-memories used, on the cognitive internal processes at play, and on the internal and external knowledge incorporated and constructed during program comprehension. Other authors, such as Minsky (1974), Rich and Waters (1990), and Soloway (1986) theorise the cognitive internal representation of knowledge through the concepts of frames, plans, and chunks. However, no existing theory of program comprehension explains well the difference between expert and novice software engineers, in particular wrt. the use of patterns.