What does “early stage” mean in “early stage technology”? Does it arise from DoD “Technology Readiness Levels“? Perhaps. With TRL treatment, we are deep into the Linear Model. The starting point is basic research, which has to be translated into applied research, and thence through systematic development to demonstration, testing, and launch as product or cruise missile. It is all very rational, thorough, even pat. One might also add often slow, expensive, wasteful, and fussy–but who are we to go up against the consensus?
In discussions of this sort, topics come in pairs. There is the part you like, and its chosen adversary. Here, one response is, well, you can accept our TRL development (which we love) or you can advocate for foolish, uninformed rush to implement stuff that might not ever work and cause huge disasters at great expense and be blamed for it all (which we hate, as do all properly reasonable people). The challenge is to change the dichotomy, not argue it. One might ask: are there any alternative developments of a TRL schema that would move more quickly to a resolution about viability? Are there ways to reposition the source of new technology so it arises within the systems in practice and basic research is drawn in to wrap general theory around stuff that already works in local situations? Is it possible that the whole TRL process has been constructed to deal with improper claiming in the context of procurement contracts, and that at heart this has little at all to do with technology development and a whole lot to do with what people say when they want money, and the people handing it out don’t rightly know what anyone is talking about. That is, perhaps TRL is an effort not to be gamed by folks trying not to be feckless.
Or, one can go with dichotomy, join in with those in love with TRL as formulated, and show that one is a true buddy and no advocate of chaos and pain. In Harrison White terms, we might say that a TRL dichotomy aims to create an “arena” where the fundamental issue is whether one is “in” or “out”. In terms of set logic, it creates a counting principle to divide folks up into groups. Politically, it aims to create a bimodal distribution with the major peak being those that accept TRL, and who are certified to obtain rewards (in the form of federal contracts) and a smaller, preferably not organized remainder who are excluded.
Overall, the usage “early stage technology” points to a genre of discourse on research that aims to rationalize inquiry with regard to how anything new gains acceptance. The emphasis is on a movement from uncertainty to certainty, of progressive validation and improvement, until something new is fit to join existing technology and worthy to replace other technology to be retired. Much as one might handle officer training. Eventually, you may become an admiral, but first we will prove you out in the lower ranks with limited responsibilities to see what you are made of.
We may go further and ask what is it that suggests technology has “stages” and whether these are inherent in the technology, or are stages of acceptance by the status quo, or by purchasing officers tasked with representing the status quo, as it seeks out things that are new, but that preserve the roles and status of the status quo. The navy does not seek out new technology that would end the practice of the navy as we know it. Only enemies of the navy might do that.
There are studies around that argue that the Linear Model arises as economists confuse early reports of NSF activity, separated into provisional categories of basic, applied, and development, into an implied chronology starting with basic–that is, a process, with inputs and outputs, nicely situated for econometric analysis for productivity and efficiency and the like. Again, it is entirely possible even if the early stage economists were feckless, that there is something to it–that folks discover, futz, get bolder, futz seriously, companies and governments get interested and get folks to futz a lot, especially on their own time to reduce actual outlays, and then right as all the really juicy core patents expire, adopt and make another pile of money, or scare the floating bejesus out of some foreign navy.
But I’m not persuaded the Linear Model means much for IP practice. (Well, it means a lot, since it gets invoked a lot, and people get fussy if you question it, but as for being a good guide, it’s up there with What Would Journey Do? Basically, the Linear Model is a weak metonymy that has service in justifying systematic university research to the public and competing for budget within federal budget allocations.
In the construction of fictions, there are two great traditions. One involves metaphor. This is pretty well known. Transfer a meaning from one context to another, as a comparison (simile) or as an assertion (common metaphor). He is like a tree (simile–in some way). He is an oak (meaning, big and strong perhaps, rather than balding, which would be an oak in winter). The other involves metonymy, which is rather less well known. Metonymy involves substitution of an instance for a class, or vice versa. This may be a part for a whole (synecdoche) or a closely related thing or attribute for the thing itself (common metonymy).
The beauty of metonymy is that it typically reports something verifiable rather than simply comparative. If I say, the White House reported that the economy is doing better, I am not saying there is a talking house, but rather that people associated with the office of the president in the White House have something to say. But there is a White House. Similarly, I can say that I’m going to visit San Jose when really I’m going Los Gatos, but more people recognize San Jose because there was a song about trying to get there and not one about Los Gatos. Since San Jose and Los Gatos are both members of the class “cities in the Bay Area” I merely pick one that’s more recognizable than others. For general work, that’s good enough.
In new ventures, the bios of the founders stand for the experience of everyone who will join the company. Metonymy. In university technology transfer annual reports, “success stories” stand for the activity of the whole portfolio over the year. What folks want you to understand is that there are many more such stories, of which this is just an illustrative selection. Folks don’t say something like, whew it was a rough year and this it, the sum total of anything we could find that is halfway successful, and we’ve spun it out as good as we can without outright lying about things, well not more than otherwise. Nope, don’t get that. The idea is, the success stories are pieces of a much bigger whole, or a whole that the office strives for, a whole in the future that will be better than the much more limited whole of now.
So we have the Linear Model as a class of activities surrounding the development of technology. There may be a handful of instances in the class, all of a kind. And yet we have little to go on that this class is well populated. A few instances stand for an implied general practice. It’s a fact, you see. But what if Linear Model practices is only a small, difficult, long, and infrequent subset of innovation practice? Then, efforts to train people in the practice of the Linear Model would be efforts to prevent competing models to become visible, obtain the expertise needed for their operation, and even gain the policy support to empower activity. That is, if one moves from Linear Model as instance to Linear Model as the only member of the class, then one’s activities are aimed at preventing any challenge that there other members to the class.
Same for university technology transfer offices aligned with AUTM’s monoculture of practice. There’s one way to practice under Bayh-Dole, and it is owned by AUTM’s Practice Manual. Cross these folks and they come down on you all snarky and threatened. It’s understandable, because they have bought into the Linear Model as sacred text, and have sold it back to administrators, government officials, scientists, and the public as a fact of life. They cannot afford to be wrong, or limited, or no longer relevant. And, in instances, the Linear Model appears to operate. But mostly, it is just a product of early economist futzing that touches on one thing out of potentially many, made into something so serious a whole professional organization is willing to stick its collective tongue to the Linear Model’s frozen surface and stay there, with nothing else, really, that can be said on the matter.
The first challenge, then, in looking at “early stage” technology is to get a grip on whether this is a workable mindset for dealing with innovation. In one sense, sure, one can make these assertions, kick up a TRL analysis and brace for the next decade of testing. But what if the technology isn’t “early stage” (meaning not labeled that)–what if it is an adoption of something already well established in another area of practice–that is, it is mature technology being repurposed? What then? Outlier! Exception! What if the technology is merely a set of partial practices and it is the full set of practices that must mature as a new platform, so the reported technology is itself neither early nor late, but is just what it is, with no TRL analysis possible because the platform itself is coming together, and at that time it may be clear that the technology, as first developed, is already as good as it is going to get, and what needs to happen is that other pieces of the platform have to be put in place. Too fluffy! Pie in the sky!
Perhaps you get the picture. It may be that the only way to get a clear sense of what is happening in practice is to abandon, at least for a time, one’s hold on terms such as “early stage” and recognize that these terms *do something*–they *create the social expectations that feed a practice model*. Call the situation something else– “complementary technology” or “way cool stuff” and you propose rather different patterns of dealing.