Unpacking Rangaswami's "Four Pillars" of Enterprise Software

Regular readers know that I spend a lot of time wrestling with the posts on JP Rangaswami's confused of calcutta blog, usually trying to clear up what I feel are serious misconceptions about the nature of work and the role of technology in the workplace. Much of what JP writes is oriented around his assertion of "Four Pillars" of enterprise software: Publishing, Search, Fulfillment, and Conversation. While I would not dispute that these are important activities in the operation of any enterprise, I am not sure I would pose them as the "pillars" of that operation, particularly when the priorities of that enterprise necessitate living in the social world of the service provider, rather than in the objective world of the manufacturing and delivery of products. Thus, I was glad to see that JP finally decided to develop a case for his Four Pillars around the more fundamental question of the nature of work in the service sector, probably in response to a recent discussion over the extent to which the use of social software could be construed as "wasting time." Here is the latest development of his argument:

Well, in most service industries, people appear to “work” by doing four things:

They look proactively for information. They search for things.

They receive information because they said they were interested in receiving that information. They subscribe to things.

They talk to each other using various forms of communication: letter, e-mail, audio, video, text, IM, blog, wiki, twitter, whatever. They are even known occasionally to talk to each other face to face without use of technology.

And they transact business as a result. Within the enterprise. In the extended enterprise and partners and supply chain. With customers.

People do all this now. But we do not have the tools to do the job well. Search is not free-form and wild-cardable and probabilistic; most of our search is very forms-driven and deterministic. Publishing or syndication is done using reams of paper producing reams of reports that no one reads. Mainly because it is done in elephant-sized chunks rather than bite-sized chunks. We need to be able to subscribe to changes in data elements rather than whole humongous lumps of data. Conversation is constrained by the difference in the technologies used. And fulfilment is often held up because we have weaknesses in the static data we use, things are not always up to date. Wrong names, misspellings, wrong addresses, wrong or missing authorisation information, poor delivery or fulfilment instructions.

That’s what Four Pillars is about. Syndication. Search. Conversation. Fulfilment.

This is all very neat, if a bit heavy on the techno-centric tosh (one of JP's favorite nouns, although I would probably go with "claptrap"); but, to try to push the metaphor in a meaningful direction, does it have the "strength of support" one expects from pillars?

Having invested a lot of my own cycles in the nature of service work and the supporting roles that technology may play there, I want to challenge the initial premise of this passage and then see where the challenge leads. In the spirit of my blog title, I should also point out that much of my challenge is based on material I have already "rehearsed" in a seminar talk I delivered at the University of California at Santa Cruz last March. That talk gave me an opportunity to present my skepticism of the recent academic "rush" to offer curricula in "service(s) science," which I had been "rehearsing" in my blog writing for some time.

The basis for my challenge is to identify the nature of service work before trying to enumerate what service providers actually do. Invoking the language of Jürgen Habermas, I wish to characterize service as a two-phase "action situation:"

1. The customer has a motivated action that needs to be performed.
2. The service provider performs that action on the customer’s behalf.

Thus, if we want to understand the nature of service work, we must first understand the nature of those motivated actions that a service provider may perform. I propose that we do this by invoking a variation on a quad chart based on the “Four Paradigms of Information Systems Development” paper that Hirschheim and Klein published in the Communications of the ACM back in the dark ages of October of 1989 (inspired by a book from the even darker ages of 1979 by Burrell and Morgan, Sociological Paradigms and Organizational Analysis):

The point behind this quad chart is that its structure tells us something about the nature of the engagement that must take place between the service provider and the customer if the service provider is to effectively perform the customer's motivated action(s). It is only in terms of the nature of that engagement that we can assess what the four elements of JP's list actually support:

1. I have already addressed the matter of search in terms of the "discontents" of the "Google paradigm;" but I think the point here is that a service provider can only be proactive about search if the search has some kind of goal. If we are taking a task-based approach to the customer's needs, then engagement with the customer is likely to lead to, at the very least, one or more hypotheses about such a goal. However, if those needs are worker-based, the emphasis is probably going to be on motives, rather than actions; and it becomes much more difficult to address those motives in terms of goals. In other words search is not much a "pillar" for half of the quad chart!
2. On the other hand the engagement will certainly always be informed by the "knowledge" that the service provider brings to the customer. However, that knowledge is less a matter of what information the provider receives and more a matter of the provider's interpretative skills in the face of information that "flows in," whether by virtue of subscription or happenstance. Without those skills it matters little what is actually "flowing" to the provider, even if it is "focused by need" or packed in "bite-sized chunks." In other words all that really matters is that the provider needs to be a first-rate reader, regardless of any benefits from publishing/syndication; and, in the spirit of Paul Ricœur, that reading skill must be applied to actions as well as texts.
3. Equipped with those interpretative skills, the provider is then in a position to organize the engagement around conversation; but it is a conversation grounded more in social processes than in the specifics of content or the media through which the content is delivered. I have already argued that JP's assumptions about conversation are impoverished in the context of this social dimension. Therefore, I shall not belabor the point further but once again acknowledge Erving Goffman as the inspiration behind my own point of view.
4. This brings us to the final point, which is the danger of viewing the engagement as nothing more than a business transaction. This is also a point I have already discussed; but, in this case, I would like to explore an additional perspective. Consider health care as a paradigm for a service offering; and consider the hypothesis that the "industrialization" of health care that we now face in the United States arises from a failure to understand the breadth of motivated actions around relations between physicians and their patients. JP's list may suffice well enough if all we are trying to address is the diagnosis of a patient's malady (goal satisfaction); but, if our ideal vision of health care is one in which the literal sense of "health maintenance" (rather than the one corrupted by the institutions that currently invoke that phrase) is one of an ongoing engagement (achieving mutual understanding and/or changing work/life practices) that takes place "in sickness and in health" (to shamelessly borrow a familiar phrase). This latter reading is not grounded in transactions and would be undermined if it were grounded in transactions.

That is probably enough for now. (Besides, I need a break for today's lunch-time recital!) My point is that the pillar metaphor can only be invoked through the connotation of support. From that point of view, JP's pillars do not support very much. Nevertheless, I believe that there is a role for the support of service through technology. I also believe that we need to do a lot more research in both the social and technical dimensions of service engagements before claiming that we understand that role and how it can be "played."