Meaninglessness of the Word Architecture

It is perhaps more foolhardy than courageous to assert that an almost universally used term is meaningless. Yet such an assertion I make here. I believe that a careful consideration of the term architecture will convince the unprejudiced judge that its various assigned meanings are so contradictory and have such uncertain reference to facts that to call the term meaningless is no more than an act of simple charity. But before taking up the argument, let us examine some statements using the term, as culled from representative IT books and articles (1) published in the last 20 years.

1) The fundamental organization of a system, embodied in its components, their relationships to each other and the environment, and the principles governing its design and evolution.
2) Information architecture spans business and technology architectures, brings them together, keeps them together, and provides the necessary rich contextual environment to solve the ubiquitous data-quality problem.
3) A formal description of a system, or a detailed plan of the system at component level to guide its implementation.
4) The structure of components, their inter-relationships, and the principles and guidelines governing their design and evolution over time.
5) By the architecture of a system, I mean the complete and detailed specification of the user interface.
6) Architecture represents the significant design decisions that shape a system, where significant is measured by cost of change.
7) Computer Architecture is the science and art of selecting and interconnecting hardware components to create computers that meet functional, performance and cost goals.
8) It is architecture, then, in the sense of how things hopefully will fit together.
9) In most successful software projects, the expert developers working on that project have a shared understanding of the system design. This shared understanding is called ‘architecture.’ This understanding includes how the system is divided into components and how the components interact through interfaces.
10) Service Oriented Architecture (SOA) is a business-centric IT architectural approach that supports integrating your business as linked, repeatable business tasks, or services.

In passing, we should note incompatibilities among the various statements: architecture is now an artifact, now a process or a roadmap; sometimes it is an entity common to all IT systems, but on the other hand “it” is different in all IT systems, even in the different layers of a single system.
Were we faced with a simple factual problem, we could demand uniform, correct statements of the authors. But our problem is no such simple one. Rather, we are faced with more subtle matters connected with the foundations of software, with the primitive conceptual frameworks that govern our thinking about “software”. I hope to show that the conceptual framework implicit in the word architecture is worse than vague, it is bad. At best, the word has functioned only as window dressing; at worst, it has probably impeded progress. It can further be shown that those research workers who are most actively engaged in determining the facts to which architecture supposedly refers have the least use for the term—surely a significant rejoinder to anyone who would suggest that the concept is a “useful fiction,” in Vaihinger’s sense (2).
In architecture we are faced with a most insidious danger in intellectual endeavors, the danger of words that are too powerful, that “explain” everything. Such enemies of thought, like all enemies, may be easier to spot if we label them. Such “explain-alls” need a name. As we borrow from the Greek to call a “cure-all” a panacea, so let us christen an “explain-all” a panchreston. The history of science is littered with the carcasses of discarded panchrestons: the Galenic humours, the Bergsonian élan vital, and the luminiferous aether Michelson and Morley tried to show with their experiment are a few cases in point. A panchreston, which “explains” all, explains nothing.

In the necessity of discarding “architecture,” IT is now confronted with a painful decision of the sort that faced its older sister science, physics, more than a century ago—the necessity of denying “common sense.” When it became clear that the “self-evident” Euclidean framework did not work for the universe in the large, there were many who would have denied the consequences of logic and experiment rather than abandon the metalinguistic presuppositions they called “common sense.” A similar crisis arose again when the “indeterminacy of nuclear electrons” was postulated. Reason finally triumphed over common sense, although it took a generation or two to do so. We would be injustifiably sanguine to expect a speedier demise of “architecture.”

References and Notes

1. I do not give the references to this quotations, believing that the works from which they are drawn should not be subjected to explicit adverse criticism. The statements are fairly representative of this class of publications.

2. H. Vaihinger, The Philisophy of ‘As If,’ C.K. Ogden, Trans. (Routledge and Kegan Paul, London, ed. 2, 1935).