The ability to write efficiently and persuasively is important in many areas of life, and especially for spreading rationalist memes and hence raising the sanity waterline. 

While there are a lot of very good and persuasive writers of both fiction and non-fiction on Less Wrong there seems to be relatively little advice on how to improve one's writing skills.

While there are a huge number of writing guides available, much like general self help they rarely reference studies on the effectiveness of the advice contained, and while some come from very successful authors, the problems of generalising from one example are well known.  

Given this, would people be willing to supply rationalist supported strategies for improving writing skills?



I've looked for previous posts on this subject, but if I have missed a previous good discussion please link to it and I will close this thread.

The most obvious piece of advice would be to engage in large amounts of writing practice, but hopefully you will be able to supply some more strategic advice than that. 



Consensus so far is that a high level of practice is very important, ideally paired with useful and continuous feedback. Otherwise a general agreement that the process is very idiosyncratic, with a few good suggestions for resources that have worked for individuals.

Ideally we'd be looking for advice that has helped a large majority of people to have tried it, if any such exists. 

(Also added links)

41 comments, sorted by Click to highlight new comments since: Today at 2:58 PM
New Comment

Write lots. You said it, but it bears repeating.

Next, get good feedback on your writing. For me that's the second most important thing.

Good feedback is feedback that tells you what effect your writing had on the reader. (This means that things like "I didn't like it" are rarely useful feedback; "I was engaged until the third paragraph, when your mention of dolphins felt unconnected to what came before" is great feedback.)

Having your writing reviewed by a group can be harrowing but is also a great way to improve fast. Diversity in the group means that most of the major improvement areas in your writing will be spotted by someone.

I don't have any studies to back that up. It's hard to find evidence-based opinions on improving one's writing. Writing is an extremely diverse activity: creative writing, persuasive writing, informative writing to name but three categories off the top of my head may have very different optimization criteria.

You may be interested in a book I'm currently reading, Anne Flaherty's The Midnight Disease. It's a neurologist's take on writing; short on advice but long on fact. It does a good job of sorting through productivity, creativity, inspiration and so on in terms of "you are a brain".

I've read a bunch of other books on writing - Wiiliams Style: towards clarity and grace (I don't remember much of it), Lamott's Bird by Bird (which I found useless), Elbow's Writing with Power (very fond of this one and the freewriting exercise in particular), Goldberg's Writing down the bones (don't remember), Richard Gabriel's Writer's workshops (best advice I've ever read on writing), etc. You could do worse than Alicorn's "saturate, distill, improvise" approach.

Writing well, it seems to me, depends on getting a great many small things just right - not on improving any particular single aspect of your writing. You need to have something interesting to say, to care a lot about it, to organize your ideas logically, to make sound arguments, to lead your reader through it without losing them at every corner. You also need to engage the reader by making your ideas relevant, to be memorable through vivid imagery, to have a sense of rhythm. You need to know when to be funny and when to be serious. And so on - and all of this needs to be tailored to the expectations of the audience and the conventions of the genre you are aiming at. A scientific paper calls for different voice than does a story for children.

"Write lots; get feedback" may be the most general advice you can get without getting into specific things you want to improve about your writing.

My best advice on learning how to write well would be to read. Read when you're on the bus. Read when you're waiting in line. People have a lot of time in a day where they are waiting. Next, as Morendil writes, is to write as often as possible. Describe the scratched, faux wood desk that your laptop/desktop is resting on, or the delicate webbing on a dried leaf and be sure to have people critique it. If you're writing fiction, have someone read over your plot first, and make sure it's someone who reads a lot and is a critical reader. Chances are they'll be able to spot plot holes or inconsistencies that you might miss. Also, take a break from something for a week and then go back to it. You'll find a slew of "I could have worded that better" -s. Most of all, be clear, concise, and passionate about what you write. If people see the author has no passion or apparent interest in what's written, they will be less inclined to read past the third or fourth paragraph. In fiction, if you are writing about a real place (for example, a character lives in Ischia, Italy) look up information about the place. Not only should you have the general attitude of the people down, but the foods, times people generally do things, local sayings, etc. (Like the majority of people living in Ischia take naps around noon because of the heat.) I am babbling, and will stop now, but I agree on the feedback. You may think you are getting your point across, but does everyone else?

Agreed, getting good feedback is very important. Something that I think may have been mentioned here before (can't remember the source) is the importance of having immediate feedback, whereas in many academic contexts feedback comes weeks after the submission of the work and rarely is specific. Edit, Spelling.

Yes, don't leave it to the professionals with a busy schedule to review your work - build a network of friends who are knowledgeable and interested enough in the subject matter to provide you with constructive and sympathetic (but lucid) observations.

LW functions almost as one such group - I'm tempted to write up a post setting out the rules in Richard Gabriel's book now that the Discussion forum can serve as a lower-pressure environment where people could post pieces specifically for the purpose of getting useful feedback on their writing.

I'm tempted to write up a post setting out the rules in Richard Gabriel's book now that the Discussion forum can serve as a lower-pressure environment where people could post pieces specifically for the purpose of getting useful feedback on their writing.

Please do so.

Will do. I'm currently on vacation so that might not be until next month.

That would be useful.

Writing, like other creative pursuits, is idiosyncratic in the extreme. Try advice if it sounds right, is cheap to try, or has a load of empirical evidence suggesting that it is helpful to the vast majority of people. Ignore it otherwise. When you find things that make prose come out of you, arrange to have more of those things. When you find things that make your prose better, arrange to do more of those things. Everything more specific than that requires personalization.

Try advice if it.... has a load of empirical evidence suggesting that it is helpful to the vast majority of people.

This ideally is what I'm looking for, do you know of anything non-trivial that fits into this category? Or a good place to start researching this. The alternative hypothesis is that an individuals conditions for producing good writing are s idiosyncratic and/or dependent on ndetectable psychological factors that nothing generalises, but that would be somewhat depressing.

[On a side note I have been extremely impressed by your writing, having devoured Luminosity and sequels with enthusiasm ]

I'm not aware of any controlled studies that have been done on the subject. People interview famous novelists about their process, but that's a weird sample, and relies on self-reports about what factors influence the writers.

Write a lot. Read a lot. Brandon Sanderson had to write five novels before he sold one. Read every day. Read Strunk and White. Learn the rules. (Yes, writing has them.) Get a giant wall calendar showing every day of the year. Set a writing goal and every day you meet it, draw a line. After a few days you have a line segment. Don't break the line! Convince yourself that reading and writing is "work" and don't let friends and family tell you otherwise. Find a room with a door. When you're writing, close it. Don't edit when you write. Move forward, forward, forward. Get to the end. If you need help with motivation, read these books.

Convince yourself that reading and writing is "work"

Be careful with this. I get lots more writing done when it's not work.

I mean it in the following sense. Writing takes time. Time that you would be spending on other pursuits (and other people!). Don't feel guilty when you've been writing. Feel guilty when you haven't been.

Most successful writers also say having set hours helps.

I have no set hours, multitask compulsively, abandon writing to do other things when I feel like it, do not experience guilt about it until and unless I am in danger of missing a deadline, and sometimes allow the unavailability of my beta readers to prevent me from writing even when I'm otherwise in the mood.

And I ship.

I'm glad your method works for you.

Ship as in deliver or ship as in romantically pairing characters? Both seem plausible on the topic of writing.

As in deliver. Shipping as in romantic pairings would have been contextually irrelevant.

Convince yourself that reading and writing is "work"

If only defining something as "work" were an effective motivating strategy for most of us!

It's probably more useful to say "Convince yourself that reading and writing is IMPORTANT."

Read Strunk and White

Or not.

The point is there are rules. Before you break them, know what they are.

No-one is doubting that rule systems for assisting writing exist, but the issue is whether they are supported by any particular evidence or are in the realm of self help book platitudes.

Strunk and White isn't self help. It's a style manual.

Apologies for the ambiguity, I meant it may have similar issues in not basing its reccomendations on hard data about what the majority of people find useful

It's very hard to talk generally, even about fiction. But if you send the same manuscript to ten different editors, the same problems come up again and again. Try not to use adverbs. Watch point-of-view. Don't use too much (or too little) description. Watch pacing. Avoid the passive voice. Show, don't tell. Use said. Etc.

In a nutshell: write a lot for helpful editors.

For two years, several hours a week, I wrote or edited for my college newspaper (The Stanford Daily) and then spent eight months working full-time in journalism internships.

I was exploring a career path but my writing skills ended up far, far better. As one commenter noted, my style tends toward "plain," "simple" and "concise." This is a reflection of my journalistic training.

My newspaper writings.

Also, if anyone wants to try this out, I can probably hook you up with one of the internships I did. It's a pretty sweet gig, you will learn a lot with a great editor in a small office. The only qualification is being reasonably intelligent. PM me.

Interesting, who would you define as a 'helpful' editor? Would it be limited to people with formal journalism skills or would someone of equal ability to yourself be beneficial for the benefit of a fresh eye?

Someone of equal or greater ability who can clearly explain their judgments.

Editors in a formal journalistic setting carry responsibility; if something is difficult to understand, wrong, or badly worded they're on the hook. Whereas a friend is more likely to say 'Oh that's nice' for fear of offending you.

I get writers block, or can't get past a simple explanation of an idea, unless I'm conversing online (usually some form of debate) in which case I can write pages and pages with no special effort.

If you have time, try Ben Franklin's method

(and tell us how it goes)

That's actually in incredibly good, rational method for learning to write. It allows the writer to study an existing model, practice, then compare against the model afterwards. One could do this with pretty much any author, journal, or paper that you enjoy, and it should be effective no matter the type of writing. Now that I think about it, I'm going to suggest it to some friends of mine who teach English classes in the area. Thanks!

I disagree with getting "useful and continuous feedback".

Getting feedback too early means you probably haven't learned the difference between drafting and editing. These ought to be separate. So while I agree asking for feedback is good, it should only be done for a finished work.

Getting feedback too early means you probably haven't learned the difference between drafting and editing. These ought to be separate. So while I agree asking for feedback is good, it should only be done for a finished work.

I disagree with you. (Although I guess you'd just say I don't know the difference between drafting and editing. I submit that I know the difference and prefer doing the first thing really well to having to do the latter thing.)

There is no right or wrong way to do this. Kurt Vonnegut is famous for revising each and every page until it was perfect and then never going back to look at them again. But his approach is the exception.

Some things that worked for me while playwriting:

Writing while talking with someone else about your writing is a good way to improve.

It's helpful to analyze your problems abstractly before getting specific. I.e. brainstorm a lot and plot your story before starting writing. However, once you've got the basic plot it seems to be a matter of taste how specific to make it before starting writing.

Even before you start writing, ask yourself what makes something funny. What makes good dialogue? Write some few-paragraph samples to try and work on specific skills.

The smarter you are and the more "foundation" (vocabulary, knowledge of the medium, etc.) you have, the more you should expect your writing to improve every time. When writing, be on the lookout for places where you could improve. Then once you've noticed, improve!

Figure out what time of day works best and schedule time to write then. I have tried evenings, afternoons, and mornings. I find I am most productive (measured by number of words on the page) in the early morning when everyone else is asleep. If you live by yourself, this may not be so crucial. Having small children changes things.

Try to get someone to pay you for your writing. This will force you to learn what people value. Think about writing as a service you provide for others.

I'm aware of resources, but as I mentioned above very few of them give any reference to their effectiveness for anyone other than the writer themselves. Ideally we'd want something that has been shown to be effective for a statistically significant number of diverse people.

How to write persuasively is a somewhat measurable skill that we might be able to find information on. How to write "well" is vague and varies a lot from person to person, and the biggest lesson I've learned reading various writers' blogs is that different things work for different people.

The rationalist answer to "how to write well" is not "Do X" but "Here are a few things that successful authors do, try several them and be ready to constantly update your opinion on what works for you until you've found a process that works."

The only constants are going to be "write a lot" and "get feedback a lot."