A new language, referred to for the purposes of this paper as the information systems business language or ISBL, is being born in the world of business. It is an artificial language (Lotman, 1990), designed to eliminate possibilities for misunderstandings in the conduct of standardised business transactions. Its primary source language is English but it incorporates information systems (IS) concepts, definitions, symbols and gestures and is therefore not a subset of English. The language has a sphere of operation restricted to interactions involving at least one autonomous IS, and is still in the very early stages of development. The development process is erratic and likely to remain so, in that the language is a by-product of pressures for rationalisation in business interactions, rather than the outcome of a conscious design activity.

Business, like other areas of human activity, is mediated by, and understood through, language. The nature of business has, however, always been such as to impose a premium on the use of literal language and the avoidance of figurative expressions (Yates, 1989). This has led over time to a reliance on strict terminology, a reliance that has been intensified by the emergence of automated systems with zero tolerance for ambiguity. In introducing systems of this type, organisations and system designers have endeavoured to eradicate the problems of meaning that can arise from different presumptions and frames of reference (Fish, 1978) by rigidly defining and controlling the context of interaction. While this has not eliminated all possibilities for misunderstandings to occur, it has for the most part substantially reduced their effects.

Most IS have a purely instrumental character in the sense that they provide support, often very sophisticated support, for a wide range of organisational activities, but cannot be construed as acting independently in a social context. Autonomous systems dealing with other autonomous entities external to an organisation are, in contrast, qualitatively different in that they effectively act as responsible agents of the organisation. These systems, of which an automatic teller machine (ATM) is perhaps the canonical example (Dos Santos and Peffers, 1995), have been assigned responsibilities for making decisions and taking actions on the basis of information given, received and interpreted in social interactions. While the activities in which they engage are repetitive and mundane, it is the principle involved that is of interest in this paper; there is nothing intrinsic that limits their sphere of operations. If meaningful interactions occur, it seems to follow that such systems must be ascribed a form of social intelligence.

Interactions with autonomous systems take the form of conversations in which progress is achieved through the turn-taking exchange of information, and the proposal in this paper is that such systems and their interlocutors can be conceptualised as speaking a language, the ISBL. While the language is very much in its formative stages, and variations in definitions and usages are still common, there is enormous pressure for standardisation of the terms used. An easily accessible example of this process in action is provided by the progressive routinisation of autonomous payment systems, enabling organisations to use common interfaces and standard payment ‘scripts’.

It is perhaps the fact that their operations are mundane that has limited the amount of theoretical interest in autonomous IS. Strategic analysts have not overlooked their competitive significance (Dos Santos and Peffers, 1995), but the systems themselves are generally not particularly complex, and have therefore been of little technical concern. The argument in this paper is, however, that they represent a social development of great potential significance, and that their emerging capacity to ‘speak’ a common language heralds the realisation of some of that potential.

The development of the ISBL is being fuelled in practice by a range of IS integration initiatives based on enterprise system (ERP) packages, electronic data interchange (EDI), government data-sharing, and business-to-business (B2B) procurement exchanges, all of which rely for their effectiveness on the implementation of standardised IS constructs including data and process definitions. It is these constructs, rather than the perceived interfaces, that enable systems to ‘talk’ both to people and to other systems. The new language is evolving at an electronic frontier where people and systems are learning how to converse meaningfully with each other, making it a kind of pidgin language (Holm, 2000; Czarniawska, 2003). Pidgin languages are compromise languages that use a restricted lexicon and a rigid syntax to facilitate trade between different cultural groups (Holm, 2000). Continuing development of the ISBL involves a compromise between English usage and the rigid prescriptions of computer systems; while English is the source for much of the terminology, definitional relations are to formal constructs and not to the flexible concepts referenced by natural language.

The general justification for the proposal in this paper resides in the explanatory power of the linguistic perspective. The ISBL concept facilitates the understanding and analysis of a range of IS-related phenomena, particularly organisational issues arising from systems integration initiatives. Several of these relate to social power, and the possibilities for autonomous systems to be used to entrench and extend existing power differentials affecting consumers, and organisations in dependent positions within major supply chains. These aspects are discussed in detail later in the paper.