Hello,
I wrote to you last time concerning feedback, which in many cases I see as the specific case of advice. Advice is difficult because it deals with generalities; the most concrete it gets is perhaps answering a few questions without looking at a text. It’s a catch-all term from that relative proximity to an essay posted on the Internet with a large, therefore meaningfully diverse, intended audience.
Feedback gives a somewhat-shaped work of clay to move around; advice asks you to imagine what that clay might be like and opine based on your own mental picture. It’s more impersonal, and yet the implicit ask is the same: make this better; if not, make me better at making it better.
Recently I have been asked to think about how I might convey some of my limited expertise in writing fiction for a small group, informally. Just as a thing to try out, not in any professed authoritative capacity, mind you. As I understand it, this person suggested it because I seem to approach the task in a relatively positive manner, I manage to write fiction on a semi-regular basis, and I do so reasonably well.
My initial reaction was slightly more nervous than I might have expected, and in the don’t-do-it rather than work-through-it direction, and so I tried to break it down as I figured out why.
There is a chest-fluttery sensation which I have when doing most things for which I am not, or do not feel myself, fully prepared. This one’s pretty easy: do some preparation, come back, and repeat. (Get stuck on “do some preparation?” I do too. Anyway,)
I never like the thought that I might be thought poorly of, even if my mammoth-brain risk assessment has little bearing on reality. Unpleasant, but ~fixable by thinking about it, more or less. (If something is still genuinely scary, maybe reevaluate or ask “why” again.)
It’s an obligation, and I feel some uncertainty around the ability to meet those— again, even if it’s unfounded.
Finally, I don’t want to waste time; consciously or unconsciously, I don’t think it’d be that useful even if I did it well.
As I think about it, I find myself breaking this up into points of failure:
Say I trust myself X% to do [show up on time, have prepared enough, get over myself, etc..]
Then, I trust the ability of my advice in this specific case to cause positive impact to be Y%*Z%, where Y = [advice works] and Z = [I can give that advice.]
Altogether, thinking about those three factors, I’m pretty comfortable with X and Z, the controllables; they're fairly high, all things considered, and in any case there’s really very little downside. I’m less sure about Z, that there’s some advice regarding writing that has the potential to cause meaningful change.
Empirically, this is poorly represented by my own experience —I’ve read a lot of helpful material— but it’s hard to puzzle out what ‘helpful’ means here, exactly. The word’s doing a lot of work.
I’ve read things, certainly, that have inspired me to write more. I’ve been pushed towards quantity by generalized statements and broad proclamations as much as positive feedback. But actually narrowing down [words I heard which made me better at writing] is very difficult to disentangle, even if it’s true.
There are complicating factors: just doing more of something, even messing around, will generally make you better at it. If you happen to be a teenager, aging a year or two and reading some more stuff will generally result in a stronger vocabulary, better grammar, and easier-to-express thoughts. With those two things stripped out, it’s difficult to determine what accounts for any improvement.
In the most obvious terms, anything that works works, i.e. if something did make me write more as its only effect, it was useful. And yet, even if I could replicate it, it doesn’t feel like what people are looking for when they ask for advice.
If something gets shoved to the back of a mental closet and somehow has some subtle effects from inside its shoebox, it’s not really tractable. It doesn’t add or multiply force with any real intentionality. Passive buffs are brilliant —I’d cast them all day, if I knew how— and some people read famous authors’ work with the intent of internalizing something.
But these kinds of attempts at absorbed knowledge are inconsistent at best and don’t seem to have any visible effects, meaning you can’t tell if they’re working. Moreover, unless you specifically keep track, anything of this nature that happens to work could easily be confused with other changes. That means it’s difficult to replicate and even harder to transmit— more so, in fact, than anything which straightforwardly fails and can be discounted.
Advice designed more specifically, to counteract particular categories of problem, fails in similar ways. Even if the information is transferred artfully, with high-fidelity, the modal result still isn’t all that great. These characteristics, in my mind, mark the opposite of ideal. There’s a lot of expectation and inferred meaning that can’t be met and isn’t there.
The obvious main problem is that people are not the same. The space of possible solutions will vary, for different recipients, as will which of those is right. Thus, the value of the same advice in practice may be wildly different. The flip side also holds true, inextricably, being that any recommendations are made by people who themselves vary.
Hunter S. Thompson, to Hume Logan (link:)
“…advice can only be a product of the man who gives it. What is truth to one may be disaster to another. I do not see life through your eyes, nor you through mine. If I were to attempt to give you specific advice, it would be too much like the blind leading the blind.”
Among other thoughts, Alexey Guzey writes:
“Even if someone shares our thinking style (so we feel like we have good rapport and understand each other well), they probably differ from us in a lot of other crucial personality aspects and thus are unlikely to understand our attitude towards life by default. Everyone (including me) tends to really underestimate how different we all are.”
These inaccurate models, even of ourselves, and our short-term preferences can lead us to worsen this problem inadvertently. If a ‘solve’ here is a good match of advice to its recipient, it’s clouded by biases in data towards things that feel good to hear.
“Suppose a lot of that stuff about bravery debates is right. That lots of the advice people give is useful for some people, but that the opposite advice is useful for other people.”
“I wonder whether everyone would be better off if they automatically reversed any tempting advice that they heard (except feedback directed at them personally). Whenever they read an inspirational figure saying “take more risks”, they interpret it as “I seem to be looking for advice telling me to take more risks; that fact itself means I am probably risk-seeking and need to be more careful”.”
Note the distinction: feedback is personal, tailored to a specific set of circumstances. If it’s writing, the text is at-hand, instead of a vague idea.
Put another way, by Agnes Callard:
“As I’m using the word “advice,” it aims to combine the impersonal and the transformative… She wants the kind of value she would get from the second, but she wants it given to her in the manner of the first… The problem here is a mismatch between form and content.”
The issue is similarly clouded by the great difficulty in quickly sharing the (likely many) complicating idiosyncrasies between both parties without a high-touch process. A prospective advice-giver can broadcast these beforehand, with caveats and clarifications. Any detailed follow-ups, however, are impossible (even if they can in theory clearly communicate the surrounding conditions given time!)
The reverse —the receiver of advice attempting to get across their situation— is by definition essentially impossible, though it’s probably even more important. It’s tough, on both sides, to pick apart the what, the why, and the how.
“Interviews with writers often touch on their writing process to try to explain how it is done; the hope of the reader is, deep down, to learn how they do the things they do and perhaps the reader can do the same thing. For the most part, the lesson I’ve taken away from such profiles is that every writer is different and there do not seem to be many generalizable practices, if indeed any of them matter…”
To this point, I’ve been half-assuming a few key things: you’re trying to give Advice, i.e. not Feedback and not Criticism. You want to address as large an audience as possible, effect the most possible positive change, and expend the least possible [effort, time.] Broadly speaking, these considerations create the problems— but removing those constraints only creates more problems. It’s all trade-offs.
Neal Stephenson, on opportunities to “interact directly with readers:”
“There is little to nothing that I can offer readers above and beyond what appears in my published writings. It follows that I should devote all my efforts to writing more material for publication, rather than spending a few minutes here, a day there, answering e-mails or going to conferences.”
He refers, above, to books— but it is my observation that the same philosophy would relate directly to giving advice (in “published writings”) versus attempts to be more personal. When you narrow scope, you lose something. This is true across all dimensions, and it is why the Advice Problem is worth solving.
Much like [insert political philosophy,] if advice-giving were implemented “correctly,” it’d do a lot of good which otherwise couldn’t be accomplished. To mix analogies horrifically, it’d eliminate significant deadweight loss.
Here is a considered half-solution to that main Advice Problem, then: do the things that the umbrella of Advice, as we’ve defined it, does well, and do as little as possible outside those bounds.
Things Advice does poorly, for reasons we’ve described:
connect the cause-and-effect dots in your applications of Advice
ensure that you carry out the recommended thoughts or behaviors
directly solve your (yes, your(!)) problem
Things that Advice does well:
thought-out, very general prescriptions
the nitty-gritty, descriptive details of authorial process
offer [algorithms/guidelines/a flowchart] for drawing conclusions from the previous points
That is, treat the bugs as features.
Advice has the benefit of front-loading time. Therefore, it can try to anticipate potential objections and streamline them. One frame for this is the enemy control ray: how could the advice you’re trying to impose be completely misinterpreted by someone who happens to have different assumptions?
Take the most generic possible form of advice —”write in the morning,” for example, or “write every day”— and break down the factors that it relies on. What makes it work, beyond “it seems to be common knowledge;” what is the hard-to-follow habit actually getting at? This is the great advantage: those questions can be asked, answered, and written out beforehand. Those answers are the good stuff— that’s what the advice actually should be.
Advice has the benefit of being a one-sided transaction. That is, one party has a heavily weighted ability to give out information, though they do have little of the knowledge on which proper prescription would rely. It is possible to freely dispense serious amounts of descriptive material, however. It’s hard to do damage with “here’s what I think works for me, and why,” and it has positive utility, at least potentially.
With a large enough sample, the need for an excellent matching system is mitigated, even perhaps obviated. It’s like the Secretary Problem, except you can keep plugging in inputs. This introduces the danger of switching things prematurely, but that’s counterbalanced by the fact that you don’t have to try anything. Thorough explorations of “here’s how I work best, probably” isn’t perfect, but it’s markedly better than nothing— and, I suspect, better than the status quo of “I don’t know, [platitude.]”
And some suggestions along the lines of this piece, a pre-mortem fault analysis, can help to put those parts together in a way that helps. It’s helped me, anyway; I’m a little less nervous now.
Best,