Industries fall into catagories by their style of working.

Some industries are cyclical.
For example, farming, high fashion, and class room education are on a yearly cycle.
Management in these industries is typified by 'dialing in' the process, 'next year we'll have more computer terminals
at enrollment' and incremental change, and protecting the system 'find a substitute teacher for Doc Bailey, he's got the flu'.

Some industries are 'campaign' oriented.
For example, chemical industries and steel mills (where the term comes from) operate the plant every day for a long time, then shut down to renew the equipment, or theatre troupes that rehearse a show, then go on the road.
Management in these industries switches between the project and cyclical styles.

Some industries are project oriented.
For example, publishing, the film industry, and the software industry have projects.
Management is about getting the project done.

But really, every one of my examples has elements of all three styles in what they actually *do*.
The farmer might decide to build a new barn, a project, and software is released on a rev cycle.
Academics plan projects to fit into interterm - a form of campaign.

I'm an engineer, now working building e learning content. I see that there's a fundamental problem.
Education has a management system set up for a cyclical industry.
E learning is fundamentally project based.
This mismatch is reflected in things like the reluctance of education to admit to needing 'outsider' skills. After all, the wise manager of a cycle tries to do what's needed with the resources at hand - because next year there will be the same requirements - every cost is marginal. For a 'pure' project, every cost is capital.
Campaign industries become wise about the two modes. The local chemical plant closes down a unit and lays off the local workers, bringing in skilled contract labor to do the rework of the plant. When the overhaul is done, the contract workers go away and they call back the local force.

Labor politics aside, there's something to learn from these styles. We need to start realizing that sometimes the best e learning experiences will come from projects that will require significant engineering and design labor.
An example would be the 'zoology' entry from my blog.

Teaching became cyclical in a day when the cost of writing was sufficiently high to make it a specialist skill. Accordingly, the only way to transmit knowledge was the teacher as performer. Now we have literate, or at least minimally media literate (as in 'can watch TV') students. Why, pray tell, are we still managing using a cyclical system?

(Labor politics again rears it's head).

Liza (who I hope will chime in here) has suggested that the issue is one of conservatism. Education transmits culture. If it doesn't get transmitted, disaster happens, so educators, like farmers, become conservative.
(and her example is wrong - hunter gatherers are much more innovative than agricultural societies)

I guess this leads into the question of why cyclical systems tend to be conservative. The answer, I think, is in this 'dialing in' management. If you do something over and over you get really good at it, but only if you do it the same way each time, only changing a few things each iteration,a form of science integrated into practical work.