On Advice

Good advice is powerful. My guess is that you could argue circles around Aristotle, not because you are cleverer than he was (though perhaps you are) but rather because you are building on millenia of accumulated ideas. You probably learned principles of logic without even noticing it that would have puzzled him for months. Concepts that took humanity ages to develop are now second nature, and it can seem baffling that it was ever hard to think otherwise. In this way, passing on and using good advice is in some ways literally the foundation of civilization.

At the same time, without meaning to discourage you (or to diminish the value of this text!) advice is very hard to both give and receive.

Giving Advice is Hard

Advice is mostly hard to give because we have different ideas of success, because we do not understand what causes success, and because we do not understand the differences between individuals’ positions.

If we have different goals, my advice to you is fruitless. I will be leading you down the wrong path, and you will feel it. Either you’ll follow the advice regardless or you’ll correctly ignore it. For this reason, when I write research advice in this blog, I’m trying to be extremely explicit about what I am aiming for and why, as well as what the challenges I think we need to tackle are, and why I think my solutions address those challenges. Specifically, I am aiming towards doing high quality and important research that helps humanity, being personally happy and healthy, being respected and valued by collaborators, and treating one’s friends and family with respect. If you aren’t interested in some of that, or if you have other objectives, then you will need to reinterpret everything I say and adapt it to be useful for you.

Once you and I have agreed on our goals, and accounted for any disagreements, it is still hard to give advice because it is hard to know what causes success. For this reason, I don’t particularly think advice from very successful people is necessarily all that useful. A Nobel prize in Chemistry, for example, makes it very likely you’re a great researcher, but it may not come alongside an explicit understanding of how to become a great researcher, or even what it was that made you great.

Moreover, there is a strong temptation to think that the secret of success is whatever you did. Sometimes you will be right, but often the result is a cargo cult unconnected to the actual causes of success. In the worst cases, quite commonly, what used to work is actively not what leads to future success; because science is about building new things on top of what has been discovered, not rediscovery.

There is also a tendency to focus one’s advice on the things that one personally put effort into, which can miss the point completely if one is already very talented at the real crucial skill. Imagine Homer, the ancient poet in the oral tradition, who never learned how to use written language. If you gave him a pen and showed him how to form the characters of his language, he would tell you that the hard part about writing an epic poem is getting it physically onto the page. For you and me, that’s the easy bit.

Of course, this doesn’t mean that unsuccessful people give better advice. Success shows, at least, some familiarity with the matter at hand. Nevertheless, I have often found that the most useful explicit knowledge about good practice comes from people to whom things do not come naturally, whose skills are hard won. Someone who is effortlessly charming might have no way to even describe what they do, let alone why it works. But someone who has had to work to learn how to speak persuasively might be able to say exactly what is going on - even if perhaps they cannot really pull it off themselves.

A friend of mine, who is one of the people I regard as best in the world at the explicit procedure for making important decisions, is, themselves, horribly indecisive. And this is no coincidence. They have thought longer and harder about how to make a decision than any reasonable person ever would. That’s why they are now an expert!

This relates to the observation that there are (at least) three ways to learn. You can be taught knowledge, you can be coached in practice, and you can emulate masters. Books are mostly good for teaching knowledge - written language is far too imprecise and static to offer coaching. To be taught (in this sense), knowledge must be explicit. Mastering the practice is not necessary for having explicit knowledge, neither is it sufficient.

As a brief digression, for this reason a good academic supervisor for coaching may be different from an academic master to emulate. A good coach must be attentive and responsive and provide rapid and frequent feedback on your actions. A master could ignore you completely and still teach you a great deal by providing a model through their own work. I personally have learned a lot by studying an academic master, David MacKay, who sadly died just before I began my PhD. I think he would have made an excellent coach too, but his work and its ability to serve as a model outlived him. Most supervisors offer a mixture of these two teaching strategies, as a coach and a master, whether they realize it or not. A supervisor who neglects either aspect of their education toolkit will not develop you fully.

Finally, advice is hard to give because individuals differ. Often, when people recommend books or movies, they are actually just telling you they liked the book. They don’t particularly make an effort to figure out what you like about books, and whether the book will meet those needs. That’s fine for books. The worst case outcome is that you waste the price of a book and a couple hours before you realize this wasn’t for you and find another book.

Unfortunately, for things like careers the stakes are much higher but nevertheless people show similar neglect when offering advice. People say “You should go into field X.” when what they mean is “I thought briefly about X this morning and it struck me as interesting.” There’s no ill will, but more senior researchers are often quite ignorant of the extent to which younger ones can dramatically fixate on some throwaway thought of theirs. They have forgotten what it was like to not know what to do with one’s life, and so the awe-inspiring seriousness of offering someone advice on their career doesn’t fill them with the trepidation it should.

There is a related risk buried in the difficulty of spotting when you are using words to mean something different from their literal meaning. It is not uncommon that people give advice which is really a hope. Consider the (reasonably common) claim that high intelligence is not that important for doing good research. I’m going to completely ignore for the moment whether that is true or not, but it is often claimed (e.g., Medawar. “Advice to a Young Scientist”. 1979. or Wilson. “Letters to a Young Scientist”. 2013.). But someone saying it might be saying it because they think it is true. They might also be saying it because they mean they think it ought to be true. They might even be saying it because they would like you to think they are the sort of people who believe it. Or perhaps they are saying it because it is useful for them if you believe it, regardless of whether it is true. It can be very difficult, as a giver of advice, to really know for yourself which of those four things you are doing. And an unscrupulous advice-giver might knowingly say things for reasons other than their truth! Certainly any receiver of advice must be on their guard to tease these things apart.

That’s why all of my writing on this topic comes with a big disclaimer. I promise you I’m trying to offer the best guidance I can. But this whole thing comes with warning lights, and they are flashing. For the main thing, please remember that I don’t know you and I don’t understand your problems. I have written the same words for every reader, so it’s your job to pick out the lessons that are there for you.

I do try to take this into account when I write about doing research, and to hedge in the right places. But I won’t always notice everywhere I ought to hedge.

For example, my approach to academia is basically a “high-trust” approach in which you play fair with folks and don’t worry too much about getting taken advantage of. My feeling is that sometimes I’ll come out poorly by doing this but that in the long haul I’ll do just fine and I’ll be much more satisfied and much more able to explore interesting collaborations. I’m not usually worried about people stealing my own ideas (though of course I still place a high value on confidentiality with my collaborators’ ideas, so that people feel comfortable sharing their work with me). In my own personal experience, I know that I have received incredible ideas from collaborators, that papers have been written by others that I might never have gotten around to writing myself, and that I have been generously included as a co-author much more than I have been ungenerously left out.

But I am really very secure, for example in terms of social and career capital. If someone did steal my ideas, I would probably be fine, and people would probably believe me if I called someone out for foul play. I suspect that unconscious bias often plays in my favour, and generally in ways that I don’t perceive. Moreover I’m surrounded by people who also have the security and privilege that lets them adopt the “high-trust” style of research. If you are not in an environment that allows this, then you may need to adapt the ideas in this book pretty extensively, and probably in ways that I cannot foresee.

Having said that, I do think that many people who think they can’t afford to have a “high-trust” style are wrong. They deny themselves opportunities in the name of stopping people from stealing their work and for very little gain. As a rule, the best way to protect your “ownership” of ideas is to make sure lots of people know what an interesting bundle of thoughts you have in your head! That means you protect your ideas much better by talking about them to anyone who will listen than by keeping them secret.

What I mean to say, by this example, is that although the strategy I describe is not always optimal, and in your situation there may be specific circumstances that mean you should heavily adapt or even reject anything I suggest, nevertheless, I do stand by what I recommend. Every PhD student I have worked with would benefit from adopting the research mindset I describe more fully.

Receiving Advice is Even Harder

The last sentence of the previous section was a test. If it made you suspicious, well done. If it didn’t, go get a cup of coffee before we continue. I’ll wait.

Could it possibly be true of all of them? How would I know? What does it even mean? Maybe they would benefit, but would they also lose out? What’s the net position?

Those are all good questions, but imagine that I could answer them satisfactorily (I can’t). Even so, it only stands to reason that I developed my ideas about how to do good research in light of my experiences watching researchers. As a result, even if it’s true that they would all benefit from the mindset I describe that does not provide independent validation. The author is probably overfitting. And who are his ‘subjects’ anyhow? What’s the sample size? How big a shift in mindset would they need to make?

Understanding and using advice is a lot like understanding and using a research paper. You must critically evaluate the evidence and argument and then form a conclusion.

If you are not asking a lot of skeptical questions while taking advice, I expect it to end poorly. Probably it won’t be that bad, you’ll just not really take it in, nothing much will change, and no harm will be done except a missed opportunity. Possibly, you’ll succeed in picking up a new approach that isn’t very suited to your situation and have a few unnecessary injuries before you spot and fix your mistake.

The trouble is, answering these questions is an awful lot of work. All advice must be evaluated, not just listened to, and must be adapted, not just followed. Especially if you come from a background of ‘hard’ science: you will find all the evidence for any course of action inadequate, but nevertheless you must do something.

Sometimes your evaluations will be wrong, and often your adaptations will miss crucial aspects. I have often found myself only really understanding the advice I was given after several years have passed, because only after making some mistakes was my conceptual framework close enough to that of my advisor that I was able to understand what they had told me. I wish that I had been wise enough to see what they meant the first time around. How much easier my life would have been! But I wasn’t ready to hear the advice at the time it would have helped me the most.

Unfortunately, I don’t believe there is a way around this. We want to become excellent, but becoming excellent is difficult, and requires some mistakes. Hopefully, you can stay awake to your results in a way that lets you catch mistakes fairly quickly.

This is easier with an observant coach who can reorient you when you go astray. It helps when your coach knows your process and has visibility over your work. For this reason I am skeptical of the wisdom of keeping secrets from your supervisor, especially regarding things you find hard. Some people feel that they can’t tell their supervisor about things they struggle with because they are scared the supervisor will think less of them and won’t give them a good reference. This is short-sighted, because it makes it harder to get their help, and probably doesn’t give your supervisor the respect they deserve. You can trick a supervisor into thinking you are unaware of your shortcomings, but you probably can’t trick them into thinking you’re a good researcher.

You can, however, trick them into thinking you are a bad researcher. This is especially easy to do if you believe it to be true of yourself. If you are open to your supervisor about your struggles, it becomes especially important to be open about your successes as well. But by being open about both, you empower a good supervisor to help you become the best researcher you can be. If you don’t have a supervisor, hopefully there will be someone else who can be your coach, who can help you spot when advice is working for you, and when it is not. Many academics are very busy, but as a rule they would not be academics if they didn’t enjoy teaching enthusiastic and promising students.

Besides, people absolutely love giving advice.~~~~

Twitter  Email