Venkatesh Narayanamurti

“So long as module improvement respects the protocols by which the module connects to other modules, module improvement can proceed independently of those other modules.
An extreme case of this is when the protocols are between different levels of the modular hierarchy and when there is richness on both sides of the protocol. When the upper side of the protocol is rich, the knowledge base on the lower side of the protocol is often referred to as a 'platform' on which knowledge modules above it can be based. In science, Newton's laws were a platform on which both celestial and terrestrial mechanics could be based. In technology, the personal computer software operating system is a platform on which a rich set of software application can be based. Moreover, when the lower side of the protocol is also rich, the shape of the knowledge network becomes hourglass-like. In the case of technological knowledge, the waist of the hourglass is a distinguished layer or protocol, with technologies underneath implementing the protocol and technologies above building on the protocol - with both sides 'screened' from each other by the protocol itself. As a result, the number of applications explodes independent of implementation details; similarly, the number of implementations explodes independent of application details. The number of software applications built on the Windows operating system is enormous; the number of hardware and software implementations of the Windows operating system is also enormous.
In other words, imagine two complex adaptive systems, one organized modularly and one not. At one moment, both might be able to exploit their environments equally and thus be equally 'adapted' to their environment. But they will evolve at vastly different rates, with the one organized modularly quickly outstripping the one not so organized. Modularity appears to be an evolved property in biology, one that is mimicked in the organization of human knowledge.”
― Venkatesh Narayanamurti, The Genesis of Technoscientific Revolutions: Rethinking the Nature and Nurture of Research

“Evaluation must be done in hindsight, after the work has been done, not for proposals for work to be done. That said, secondary factors must carry some weigh, because research results depend too much on historical contingency and luck. As articulated by Ralph Bown, vice president of research at Bell Labs from 1951 to 1955:
'A conviction on the part of employees that meritorious performance will be honestly appraised and adequately rewarded is a necessary ingredient of their loyalty. This appraisal, to be fair and convincing, must be based on the individual's performance and capabilities rather than wholly on the direct value of his results. A system which rewards only those lucky enough to strike an idea which pays off handsomely will not have the cooperative teamwork needed for vitality of the enterprise as a whole.”
― Venkatesh Narayanamurti, The Genesis of Technoscientific Revolutions: Rethinking the Nature and Nurture of Research