Open source software development—Just another case of collective invention?

Abstract

Do open source software (OSS) projects represent a new innovation model? Under what conditions can it be employed in other contexts? “Collective invention” regimes usually ended when a dominant design emerged. This is not the case with OSS. Firstly, the OSS community developed the institutional innovation of OSS licenses enabling OSS software to survive as a common property. Secondly, these licenses are mainly enforced by pro-socially motivated contributors. We characterize the conditions under which OSS communities develop and sustain pro-social commitments. We point out the vulnerability of these conditions to developments in patent legislation.

Introduction

Open source software projects are examples of an innovation model, which relies neither on the price system nor on formal hierarchies nor on alliance agreements. Does the production of open source software (OSS) represent a new model, and under what conditions can it be used in other contexts? OSS is a term for software published under licenses that do not give any private intellectual property rights to the developers. To a large extent, contributions to these projects are made by voluntary, non-paid members of the OSS community. These members invest their own private resources into a public good.

A look at history shows that OSS production seems to be similar to other examples in the past and present that can be summarized by the term “collective invention” (Allen, 1983). In the second half of the 19th century, iron-making companies in Britain's Cleveland district willingly shared their innovations in blast furnace design. Other examples include the enhancement of steam engine design after 1800 (Nuvolari, 2002) and the search for a dominant design in the flat panel display industry (Spencer, 2003). According to Von Hippel and von Krogh (2003), OSS represents a novel innovation model, which they call the “private-collective” model. It shares many similarities with the “collective invention” model. However, most collective invention regimes did not survive after the development of a dominant design (e.g. Meyer, 2002), while OSS is still alive and well in such a phase of the innovation process.

This paper addresses two research questions: (1) Why did “collective invention” usually not survive after the development of a dominant design? (2) What are the conditions that enabled the OSS collective invention to do just that? Answering these questions can help determine under what conditions “collective invention” or the “private-collective” innovation model can exist and be employed in other contexts than OSS (see e.g. Benkler, 2002, Benkler, 2004, Von Hippel and von Krogh, 2003).

In Section 2, we give a short overview of the main characteristics of OSS. In Section 3, we present three other examples of collective invention. In Section 4, we explore what the different examples have in common. We discuss why collective invention normally does not survive the whole innovation process. In Section 5, we compare these findings with the OSS innovation process. We argue that OSS licenses, which are copyright-based, enable the collective invention mode to survive in a copyrighted world. In order to gain a better understanding of how OSS licenses can help with obtaining that goal, we describe the mixture of selective benefits that coexist in OSS compared to other collective invention cases in Section 6. We then ask what conditions exist preventing OSS making the changeover from a “private-collective” innovation model to a “private investment” model. We argue that the characteristics of OSS that are often identified as crucial, namely to be (a) an information product, (b) a user innovation that is (c) the product of a highly modular design, cannot fully explain the continued survival of the collective invention model. But they enable what we call a low-cost situation, which facilitates contributions to public goods. We argue that only the interplay of low-cost situations, OSS licenses and conditions that foster and maintain pro-social intrinsic motivation can explain the success of the OSS innovation model. We conclude with some remarks concerning the vulnerability of this model to developments in patent legislation.