I had the plea­sure (and honor) of vis­it­ing Jason Moore’s lab at Dart­mouth ear­lier this week, and giv­ing a lit­tle sem­i­nar ver­sion of some­thing big I’ve been work­ing on for the last a few months. More about that project in a few days; the visit helped clar­ify a num­ber of open ques­tions and focus atten­tion where it was needed.

This was my first “real” visit to an aca­d­e­mic envi­ron­ment in a few years—the sort where I’m not just lurk­ing in the back­ground and hang­ing out with my tenure-​​track friends. Indeed, the last time I did some­thing like this I think it was my 2008 visit to Nic McPhee at the Uni­ver­sity of Min­nesota at Mor­ris. Like Jason, Nic was also nice and help­ful, but UM Mor­ris a qual­i­ta­tively dif­fer­ent aca­d­e­mic cul­ture from that of the med­ical school at Dart­mouth. Both times I vis­ited mainly to observe the local work cul­tures, espe­cially look­ing at the col­lab­o­ra­tive net­work that con­nects stu­dents, fac­ulty and staff—within and between their respec­tive labs, depart­ments, dis­ci­plines and institutions.

I’ve been build­ing a cat­a­log of cul­tural and insti­tu­tional rou­tines and obsta­cles that side-track—and (often per­ma­nently) delay—potentially valu­able projects that could oth­er­wise be explored quickly. The same old ques­tion I always ask, more or less: What do you wish you had more resources to pursue?

Recently I’ve found a use­ful way to explore these rou­tines and obsta­cles is to dis­cuss lit­tle coun­ter­fac­tual sce­nar­ios and see what bub­bles to the sur­face. It can be an inter­est­ing way to sur­face trans­gres­sive behav­ior with­out actu­ally, you know, try­ing it out in real life.

Here’s a vari­ant that came to me as I stared out an air­plane win­dow recently:

Sup­pose a highly-​​respected but soon-​​to-​​retire researcher in Com­pu­ta­tional Phys­i­ol­ogy vis­its the salient depart­ment at Large Ivy Uni­ver­sity to give a sem­i­nar. As one comes to expect from a late-​​career lumi­nary, her talk tends a bit towards the philo­soph­i­cal, but it brings up a num­ber of inter­dis­ci­pli­nary ques­tions and uncon­ven­tional approaches to the con­struc­tion, use and study of Com­pu­ta­tional Phys­i­o­log­i­cal sys­tems. There’s a lot to think about, and a lot of mate­r­ial that most main­stream col­leagues just don’t run into very often.

After her sem­i­nar, she spends a day or two vis­it­ing her Host’s lab and a few of his col­le­gial LIU labs, chat­ting with staff, stu­dents, junior fac­ulty, and their var­i­ous Prin­ci­pal Inves­ti­ga­tors about their ongo­ing research and tech­nol­ogy, and com­par­ing notes on the inter­est­ing things that folks in other insti­tu­tions and dis­ci­plines have been doing.

As it devel­ops, she takes an inter­est in one of the ideas a grad­u­ate stu­dent brings up in pass­ing. The idea isn’t a part of the student’s the­sis research, nor is it even salient to the funded projects in any of the LIU Comp Phys labs. But it’s a good idea, and she decides it would be fas­ci­nat­ing to see how it would play out, and (even bet­ter) it’s a purely com­pu­ta­tional project that the vis­it­ing scholar real­izes could be done in a few weeks… by an agile team of soft­ware devel­op­ers. It wouldn’t need a grant or even a long plan­ning or pro­posal process to see what happens.

Nei­ther LIU nor the visitor’s home insti­tu­tion has any­thing like an “agile team of soft­ware devel­op­ers” as a component—hah! Not even a lit­tle bit. But in her increas­ing time spent “out in the world”, the vis­i­tor has actu­ally run into folks who have worked in those envi­ron­ments, and started to see the point of the var­i­ous “agile val­ues and practices”—at least as a kind of Utopian ideal.

Mind you, this idea isn’t any­thing com­mer­cial. But it’s a damned inter­est­ing project, and to be frank it would be a pity to see it delayed until the stu­dent grad­u­ates, and fin­ishes her post-doc(s), and gets done with tenure track, and so on and on.…

So the vis­i­tor chats online with a few peo­ple she knows, and they agree the project as sketched is a fea­si­ble way to spend about a month of work. Obvi­ously the stu­dent should have the lion’s share of aca­d­e­mic (and other!) credit if it goes for­ward. But the agile folks she chats with remind her that the point of the “one team” prac­tice is that the stu­dent prob­a­bly needs to be co-​​located with the team doing the work with her.

Alas, the stu­dent has a the­sis com­mit­tee meet­ing com­ing up shortly. She’s been asked by her com­mit­tee to work over the draft bib­li­og­ra­phy and bring it more in line with the stan­dards expected in the high-​​impact jour­nals in the field: get rid of those weird ref­er­ences from graph the­ory and ecol­ogy papers and add more from the mod­ern Comp Phys lit­er­a­ture, for example.

Noth­ing like this project has ever been in any of the Comp Phys jour­nals. It may not even catch on in the com­mu­nity, com­pared with the more obvi­ously recep­tive audi­ence over in Arti­fi­cial Men­ta­tion. But the AM folks have never even con­sid­ered Comp Phys as a domain where their stuff might be use­ful. It’s a blue-​​sky project, in that sense.

What has to hap­pen to get this work done? Does the stu­dent leave for a month? Does every­body wait until “it’s safe”? Does the student’s advi­sor col­lab­o­rate with the vis­i­tor on a grant, and use the funds to (even­tu­ally) fund an in-​​house (and almost cer­tainly inag­ile) devel­op­ment project that will take sev­eral years to do what might hap­pen in a month under other circumstances?

Who gets credit? The vis­i­tor wants the stu­dent to get essen­tially all of it. Does the student’s advi­sor get some? Under what circumstances?

Who gives per­mis­sion? Who needs to give per­mis­sion? The stu­dent should be work­ing on her the­sis. The advi­sor should be see­ing to his student’s pro­fes­sional track. And so on.

Who is a risk? What sort of risk?