From Playing to Understanding: The Transformative Potential of Discourse versus Syntax in Learning to Program

Michael Thomas Flanagan1 and Jan Smith2

Presented at the Threshold Concepts within the Disciplines Symposium
Glasgow, UK
30 August—1 September 2006


First year electronic engineering students are no exception to their contemporaries in their addiction to computer games. However, their mastery of these lucrative products of object-oriented programming (OOP) does not readily translate into an understanding of OOP when presented as a formal first year course. One of the fascinations of teaching programming is that, whilst many students learn to program without apparent difficulty, a significant proportion finds the activity extremely troublesome. This observation may be compounded for students of electronic engineering, where threshold concepts (Meyer and Land, 2003; 2005) may be ‘nested’ in the curriculum. Such potential thresholds may lie in the concepts of OOP, in the exemplifiers dictated by electronic engineering syllabi or in the linguistics of a computer language itself. Implications for teaching and curriculum redesign vary significantly across this spectrum. In this context, students’ problems appear to arise from two sources: firstly, the form of the programming language which, to paraphrase Andersen (1990), parasitises English but cannot be read as English, an overwhelming threshold conception, or secondly, more localised threshold concepts inherent in OOP itself, such as abstract classes and interfaces. Consequently we have adopted a three-fold schema to discuss these potentially troublesome concept:

This talk will discuss these three streams, in terms of the Meyer and Land threshold concept, drawing on the experience of teaching first and second year Java and C language programming courses and on a second year course on object-oriented programming for systems and control engineering. The locally challenged and conceptually challenged streams will be discussed primarily in terms of a compounding of local examples of troublesome concepts in programming with those in the electronics and applied physics underpinning the programming exercises. A linguistic approach, borrowing concepts such as markedness, will be applied to the analysis of the operationally challenged stream.

References for the abstract

  1. Michael Thomas Flanagan, Department of Electronic and Electrical Engineering, Roberts Building, University College London (UCL), Torrington Place, London WC1E 7JE (

  2. Jan Smith, Centre for Academic Practice and Learning Enhancement, University of Strathclyde, Graham Hills Building, 50 George Street, Glasgow G1 1QE (