Clearly task analysis is a large, important and confusing area. . . . Acronyms proliferate and are easily confused. The papers describing the techniques are scattered over a number of journals, books and technical reports. Often the techniques have only been used by their authors. Terms are used by different writers to mean different things. A small sample of the terms used includes: plans, goals, methods, operations, actions, objects, procedures, tasks, subtasks and projects. Central, of course, to task analysis is the notion of a "task". However, some authors argue that a task is device-independent ( Bösser, 1987) and others that it is device-dependent and it is the goal which is device independent (e.g., Shepherd, 1989).
-- Benyon ( 1992a, p. 105)
Theories provide patterns within which data appear intelligible. They constitute a conceptual Gestalt. A theory is not pieced together from observed phenomena; it is rather what makes it possible to observe phenomena as being of a certain sort, and as related to other phenomena.
-- Hanson ( 1958, p. 90)
Because this book is describing a relatively new paradigm for work analysis, it is important that we define the key terms that we use. In doing so, we hope to provide some conceptual coherence, consistency, and clarity to the sometimes-impressionistic fields of human factors, human-computer interaction, and cognitive engineering. Ideally, all of our terms should be defined formally. Unfortunately, our thinking is not mature enough yet to achieve this ambitious goal, so the terms we use have instead been defined in a less formal, although hopefully still clear, manner. Despite this less-than-ideal level of rigor, we have tried to be systematic in our use of terms throughout the book. Eventually, we hope to work toward a more formal set of definitions (see R. A. Miller, 1982, for an exemplary initial effort). Consequently, this Glossary should be considered more as a snapshot of work in progress than as a definitive, stable product.
Our intent is well captured, and in fact inspired, by the landmark effort of Dewey and Bentley ( 1949) to develop an efficient basis for scientific inquiry (see also Bentley, 1954). The premises of their work were as follows:
Science uses its technical names efficiently. Such names serve to mark off certain portions of the scientific subjectmatter [sic] as provisionally acceptable, thereby freeing the [scientist's] attention for closer consideration of other portions that remain problematic. The efficiency lies in the ability given the [scientist] to hold such names steady--