2022-01-26

Most useful advice my Ph.D. advisor gave me

Every Ph.D. advisor becomes known to their students for a few idiosyncratic phrases. This comes as no surprise, given the guild-like master/apprentice, mentor/mentee relationship of an advisor to his grad students. It's inevitable that the grad student pick up at least a few phrases or philosophies from his advisor after spending 5-6 years researching together. (This is the average length of time to complete a Ph.D. in the core sciences, according to my experience and observations.) I could recount dozens of interesting things my advisor said-- ranging from curious and anecdotal to irritating (let's be honest) and insightful; but without a doubt the advice I found most useful was:

Read carefully; take baby steps.

I'd usually hear this phrase in a conversation like this: 

Advisor: "We need to do [some task you've never done]; do you know how to do it?"

Me: "No, I've never done [that thing] before, but I found the user's manual for it online."

Advisor: "Ok... I'll give you a week to see if you can figure it out on your own. Read carefully; take baby steps."

Any mature adult ought to be able to read.* That includes user's manuals and handbooks.** My advisor was both encouraging and admonishing: you are capable of reading the manual and figuring this out on your own; and you must be able to do that, too. Reading something to check it off a to-do list is quite different than reading something to gain new knowledge you'll have to use without additional help. That's the "read carefully" part. Sometimes I have to read the same thing a dozen times until it really hits home. Now, there is such a thing as poor writing, and it may be that one can never learn what he hopes to from learn if his only [manual] is written poorly. But I've found that a more valuable assumption is that the [document/manual/literature] is of at least sufficient quality, so the responsibility for understanding it well enough for action rests on the reader (me).

"Take baby steps" is all about making incremental progress. My advisor was encouraging me not to get in such a rush to get [some result] that I would ended up floundering in non-productivity from hasty and poorly-understood [tool] use. In the long run, I believe it's actually faster to proceed through a series of well-planned and tiny but incrementally compounding steps toward a larger goal. (This philosophy also matches with the "aim low, but aim up" comment I posted about Peterson's 12 Rules for Life.) As far as I'm concerned, I'm not sure it's possible to take too small of baby steps. I feel like grad school helped me hone my ability to analyze a problem or vision and break it down into realistic and manageable goals and individual "baby step" tasks. This skill has been invaluable in all areas of my life.

Even though I'm no longer in grad school, I still hear these words every time I start to try to figure out something new or tricky.

Another good phase everyone in my lab probably heard on a weekly basis was:

Protect some time.

This one was usually in a context like, "We need to get [some important thing] wrapped up this week; so make sure you protect some time to work on it for at least a few hours each day." In one sense, he was simply asking us to prioritize a certain task to make sure it got done. But much more than that, I took this as a reminder that one must quite literally protect his time-- defend it against hostile time-invaders like trivia, distractions, busy-work, and even other things that do need to get done but are lower priority. To this end, I found the book Deep Work to be invaluable in helping me find strategies to put this advice into practice.

Here's one that didn't come from my advisor, but did come from the advisor of one of my grad school friends:

Expect it to take pi-times longer than anticipated.

A.k.a. "The rule of pi." The basic idea is that, when you're asked to estimate the amount of time a given task will require, take the amount of you would reasonably (i.e. be realistic, not over-zealous) expect someone with your expertise to require, multiply that number by pi, and that will be a more reasonable estimate of the time required. 

I don't know whether this rule was initially intended merely for comedic effect, but I have definitely experienced task competition times of something around a 3x (or greater) my estimated time. And so far it's holding true for professional and personal tasks alike!

Footnotes:

* I mean really read. I mean read, and then be able to actually use what you've read; be able to teach it to someone else; have expanded knowledge because of what you've read. Letting your eyes pass over all the words on a page is easy; reading and comprehending and retaining is the tricky part. I've often joked that I learned to read when I was in grad school, or that grad school (in the core sciences) is basically just an exercise in independent and proficient reading.

** Even worse if you're in computational chemistry: many manuals you want are for open source special-use codes written or maintained by small groups or single individuals. If you're lucky enough that they decided to write a manual for their code, it probably wasn't thoroughly reviewed, it definitely has typos, it may not have actually been tested for accurate descriptions of the program's input/output, it certainly hasn't been updated even though the code has