Defending generalists

Seth Godin is one of my favorite writers, but I have to take exception to his latest post called We specialize in everything:

When choice is limited, I want a generalist. When selection is difficult, a jack of all trades is just fine.

But whenever possible, please bring me a brilliant specialist.

If you’re shaking your head in agreement with this obvious point, then the question is: tell me again why you’re a generalist?

He later added a coda suggesting the idea of specializing in being a generalist, but sticks to his guns that “My point is that you never call on these people [generalists] when there’s a better specialist available.”

As somebody who has branded himself as an unrepentant generalist, I have to respond from my admittedly biased viewpoint.

I actually agree to some extent with Godin’s point. As The Only Sustainable Edge points out, specialization drives greater achievement in a given field, as monomaniacs achieve a level of focus that dabblers can not. Specialization also implies that only those who are truly passionate about a field will commit to the field and become the best in the world at what they do.

I think the flaw in Godin’s argument is revealed in his second paragraph: “If I need an animator, I can find the world’s best animator.” Here’s the subtle point: how do you know that you need an animator? That seems like a trivial question, but it gets to the heart of why generalists matter. Once you have defined the problem, and scoped it, and figured out exactly what skill set you need to solve your problem, then of course you’d hire the best person you can find with that skill set.

Specialists only know how to attack problems in one way – that’s part of specializing. To be the best at what they do, they have to ignore other ways of approaching the world and shut out other perspectives. A specialist is the proverbial hammer treating every problem as a nail.

So when you have a problem, how do you determine which specialist to use? Each specialist will tell you their skill set is the right one to solve the problem, because if they didn’t believe in the power of their specialization, they wouldn’t be a specialist. You need a generalist, somebody who can evaluate the problem from multiple perspectives. and who can ensure that the specialists picked will fix the real problem rather than a symptom.

The other absolutely vital role for generalists is in communication. Specialists see the world from their perspective, so for them to communicate with other specialists requires a generalist who knows enough of each specialization and its jargon to be able to translate between the worlds. This is a role that I have been very successful in filling at all of my different companies, especially on the interdisciplinary team of CellKey, where we had physicists, biologists, engineers and software developers all working together on the same product. Without a generalist, you have specialists talking past each other, and their effort is wasted because you can’t get them all working together and speaking the same language.

Maybe this is what Godin meant when he suggested one could specialize as a generalist, but I think that his post overestimates the value of skill alone, and underestimates the social difficulties of selecting and aligning specialists. The problems of language alignment and of picking the right team of specialists are where generalists provide value in a way that specialists can’t, precisely because they’re specialists.

7 thoughts on “Defending generalists

  1. Ooo, well said.

    I think there’s another flaw in his argument touched on by exactly the same quote: just because he needs and can find the world’s best animator doesn’t mean he can afford the world’s best animator.

    For tasks that are central and repeated to your enterprise, absolutely you want specialists. But for things out on the periphery of whatever you’re doing, if resources are limited (and when aren’t they?) I think you will be better off getting hold of a couple competent generalists to manage your long tail of miscellaneous tasks than you will trying to get hold of a fleet of highly-trained specialists, each of whom will tackle a single problem and then go home.

  2. There’s also the implicit acknowledgment that true specialists can’t rise above a certain level – that they’re called in to perform their functions. One must be much more of a generalist to manage the organization.

    You’re looking to manage and facilitate communication, and as such are an unrepentant generalist.

  3. In defending the generalist, I think you may have underserved the specialist. A specialist may understand that a certain tool is awesome for the job, without believing that it is the right tool for every situation, which you posit.

    For instance, people are a big fan of php for webapps and can be quite, er, enthusiastic when speaking about them, yet I would assume many/most are clear on knowing that php is not the right solution for an embedded system.

    I agree with the general gist of what you are saying here and that Godin’s argument is problematic, but can’t go all the way down the road with you on your assessment of specialists.

  4. I agree in spades that generalists are valuable and useful, because I basically make a living being one.

    The problem is that the building industry is too splintered and segmented, and nobody acts effectively as an integrator. The the structural engineer, civil engineer, landscape designer, mechanical (plumbing and HVAC) engineer, electrical engineer all work on the project in their own corners, and the architect waves his or her hands, and says, “OK guys, now be sure to coordinate with each other!” But instead, a structural beam ends up where a pipe needs to run, the guy in the field cuts a big hole through the beam, and the building inspector comes by and shuts down the job (worst case scenario there). Or the architect draws up fancy plans, but doesn’t leave any room for the cooling towers or the ductwork in the building.

    I know–and use–information on topics as diverse as weather patterns of North America, paint chemistry, HVAC equipment and controls, simultaneous heat and moisture transfer in building assemblies, construction cost data, wood behavior, and what’s on which shelf at Home Depot. It therefore puts in me in a situation where I can answer questions that the specialist cannot.

    For instance, that job that I did a few months ago–the sewer gas smell that nobody could trace down. It showed up periodically in the restaurant of a two-story building, but they couldn’t figure out what was going on. The plumber came by, looked all over, tried a few things, and didn’t fix it. The HVAC guy came by and did the same. I ended up figuring out that it was an interaction between the HVAC system (failed makeup air fan for the kitchen exhaust hood, which made the restaurant kitchen suck air from the rest of the building), and a failed sewer pipe (inside the suspended ceiling of the office of the pet store next door) that pushed out sewer gas odors whenever somebody flushed the toilet in the furniture store on the second floor. It took me two trips, but I managed to nail it in the end. A specialist really doesn’t have the tools to solve a problem like that.

    Overall, if it is a problem that requires comprehensive knowledge–like what I do, and what you used to do back at Signature–a generalist is needed to bridge the gaps.

  5. Huh… I just read Godin’s actual post, and I can’t say that I disagree with his position that being a generalist is a specialized field in itself. Or his argument that if you have the choice available, a specialist is what you will often choose. To that, I would add the caveat that it requires a situation where you understand it well enough to know that a specialist will be able to get their hands around it. Unlike the clear-cut examples that he uses (“I need an animator”), I can imagine many cases where you think you’ll need X, but it turns out that having knowledge in fields Y and Z would really help solve the problem.

Leave a Reply

Your email address will not be published. Required fields are marked *