Skip to content

“End of novel. Beginning of job.”: That point at which you make the decision to stop thinking and start finishing

From a book review by the great critic John Clute:


And here’s the part I want to focus on: “End of novel. Beginning of job.

I’ve been thinking about that line a lot recently. (I read the above review when it came out, decades ago, and I never forgot it. I was able to find it reprinted in a book via google search.)

Just about every project we work on has that “crisis decision nexus.” When do you stop playing hard—when do you switch into garbage time? When do you decide that you’ve learned enough and it’s time to write things up and tie up loose ends?

Not on page 11 of a 200-page novel, I’d hope—but at some point it happens. It’s a rare project where we go all-out from beginning to end.

I think this comes up a lot in science, and not just with my own projects either.

We’ve spent a lot of time in this space discussing junk science—beauty and sex ratio, ovulation and voting, a zillion crappy regression discontinuity studies, and that’s all before getting to cases of researchers such as the gremlins guy or the monkey guy or the pizzagate guy or the sleep guy or the criminology guy who seem to have trouble even just reporting their data—and I’ve always said that I suspect that all these researchers are believers, that even the hackiest hacks among them think their theories are true and think that their experiments or surveys or made-up experiments or made-up surveys or whatever, are steps toward that truth.

But . . . in every project, for them as for me, there’s that point where you make the decision to pack it in, to stop the sightseeing and return to base camp.

You know how, when you go on a trip, the first day—the first hour—goes exquisitely slowly, but then, as things go on, entire days blur together?

The usual right thing to do, I think, is not to try to push it beyond your inclinations. When it’s time to stop thinking hard, it’s time to stop thinking hard. No, the right thing to do is to wrap it up quickly.

To put it another way: In Clute’s telling, the author of that book had the major decision point on page 11, and he decided there and then to give up. Then the author filled the next 180 pages or whatever with insincere fluff.

But what if he had ended that book on page 12? That wouldn’t have been so bad, right? It would’ve been a short story with an interesting premise. Sure, it wouldn’t’ve been everything it could’ve been—but we’re already conditioning on the author not wanting to put in the work to do it right. Given that he didn’t want to put in the work, it was time for him to recognize that he was in garbage time and write it up.

Let’s all aspire to do this. Once you realize you’ve turned around and you’re on the way home, we should just teleport all the way back. We can the crap and give you what we’ve done.

P.S. Also this from Clute’s review: “What remains is space opera, subtly dishonored by pretense.” That describes a lot of published science too.


  1. roger koenker says:

    This sentiment seems to run counter to the spirit of relentless model checking that is expressed elsewhere here.
    Peter Galenson’s How do experiments end? provides some interesting case studies from physics.

    • Andrew says:


      It’s always better to do more—if you’re ready to do more. My point in the above post is that at some point you will lose the time or willingness to put more effort into a project, and, given that this will happen, it’s good to recognize this and not kid yourself. There’s a separate issue in science, which is that even if you’ve stopped working on a problem, other people can still pick up the torch and work on it later. So it’s good to leave your tools at the worksite, as it were, to make it easier for others to continue the work.

      • roger koenker says:

        Sure, this reminds me of a comment that Joe Doob made in an interview somewhere that at some point one’s own work becomes so repugnant that it becomes time to abandon it to the attention of others.

  2. Adede says:

    Reminds me of how someone (Krugman) described economics as a very boring subgenre of science fiction.

    • Andre says:

      Not following.

      Most jobs I’ve worked there’s no pack it in and write up what you’ve done. There’s no opportunity for write ups. In fact – I’ve been scolded by employers and associates of this community by thinking that “learning,” is a result, alluding to you saying, “you’ve learned enough.”

      If I’m payed to do a job, I’m payed to finish what I’ve done or I’m expected to give a refund and fired. I’m payed for my labor, which is usually code, not the write up. I do most things solo. There’s no team I can bounce my work to.

      I don’t really have the privilege of deciding when I’m ready to pack it in, since if I don’t finish, I’m punished. I think this might be better advice for people who are salaried to do research. I’m sure you’ve found this helpful for you, but it’s not at all for me.

      I’d be homeless if I decided to pack up and go home whenever a job needs more effort.

  3. Jukka says:

    Yes: it is a crazy world.

    Often, there are two types of stupidity. There is the subjective idiocy and fallibility (“Straussian reading”, etc.). Then there is the collective idiocy; as the old saying goes, stupidity grows in a crowd. What is sociologically interesting: as with the current plague, stupidity seems to be contagious; also intelligent people may catch it.

  4. Historian Jacques Barzun said (approximately) that in academia the chief cause of anguish is others’ works. So combined with Doob: let your repugnance become someone else’s.

Leave a Reply