This is great advice. I wish I had read this in 2021. Here's something new I learned from this post that I've been thinking. I always thought the point of the introduction was to sew tension into the existing literature and then resolve it. The rest of the paper then just fills in the details. Upon reading this, I realize that a lot of my favorite papers don't do that. After introducing the tension, the author says just enough to give you the impression they are going to resolve it, and they give you some sense of what the ingredients of the resolution will be, but you don't really know how they're going to come together until later in the paper. It's hard to pull off in much the same way jazz is hard to pull off.
You raise a more general point, which is that we sometimes don’t appreciate what our favorite papers are actually doing. When I started studying my favorite intros, I was *very* surprised to see how slow and restrained they were.
The first sentence of Jake Nebel’s “The Sum of Well-Being” isn’t about sums. It’s not about measure theory. It is literally just this:
“A person’s well-being is how good things are for that person.”
I didn't notice when first reading it how little he tips his hand in the introduction. This post prompted me to reflect on it. He introduces the conventional wisdom on encapsulation, he tells you he's going to show that it doesn't explain the thing everyone thinks it explains, but he doesn't really tell you what the key move is going to be that undermines the conventional wisdom. Still, he gives you the impression he's issuing a credible threat.
You mention that a 6000 vs 8000 word article is more likely to be accepted, ceteris paribus. But do you think that it’s a factor thats worth strategizing around? Should people aim to send 6000 words instead of 8000 to journals with an 8k suggested limit?
I had heard that going over the word limit incurs a penalty but this is the first time that I’ve heard that going under it gives you a boost. So I am not sure how to weigh it.
Of course you should eliminate unnecessary fluff etc from all your papers, but most papers seem to have some extra “bloat”, so my impression is that the “bloat” might even be considered a part of the academic style. So there could be a tradeoff between style and reducing word counts in terms of probability of acceptance.
I think short papers have better odds, partly because journals like brevity, and partly because short papers are more likely to get a quick decision.
If your paper is 6k words, reviewers are much more likely to accept the job, and they're also going to do it faster. I really can't overstate how important it is to get fast decisions -- it helps you get more feedback and send to more journals!
Also, in my experience, short papers are easier to do R&Rs for, since you have more room to add things.
This is excellent advice, and a real service to the profession. Thanks, Daniel!
One small point to add. I myself have followed the advice of 'publish on topics people you respect have been publishing on.' But philosophical temperaments differ, and I think some people are at their best forging out into (relatively) new territory. That can be a rough road, but it can also produce some beautiful and original work. It would be a shame to discourage overly much.
When forging into new territory, you have less guidance on what people are interested in from looking at the literature at what people have already been paying attention to. So for this sort of work, it is particularly valuable to present it as talks, talk over it with many people at the bar, etc., to get a sense of what are the things people find fascinating about the new topic and what are the ways you can motivate people to care about it even though they haven't thought about it already.
The advantage of publishing on topics people have already been publishing on is that you get a shortcut to getting an audience to care, because you can already see how existing papers on this topic get audiences to care. (And the existing papers may have even already created an established audience for another paper on the topic.)
I think these two intros make for a nice contrast. Kenny cuts right to the point, whereas Jake takes his time setting the scene. If you're a PhD student reading this, it might be worth pondering why these papers are (successfully!) framed so differently.
Last year, my advisee was having trouble writing an intro, so I showed her the Grice & Barnett section of this guide. Then, to demonstrate the value of rewriting, I shared the painstakingly rewritten intro of my latest paper along with a version from the graveyard.
To my horror, she told me that the new version was "Grice," and the old one was more "Barnett." And she was right!
One more thing! Sometimes I find myself graverobbing old papers for material to start new ones. The seed of "Caring Who" was a page cut from "Each Counts for One."
I don't have a graveyard - when I delete stuff it goes into the void forever. But I have deleted entire sections after realizing I can't flesh them out as much as they deserve within the paper's constraints, and resurrected them as full-fledged pieces of their own. In one case, I deleted two sections from the same paper and used them both in a new one.
What you say is probably quite right, but I rather hope people don’t follow it. And I doubly hope reviewers don’t take it on board. And that’s because many of my favourite papers would do terribly by these lights, e.g. “Killing, Letting Die and the Trolley Problem.” (How sad it would have been if Judy had received and followed this advice way-back-when.)
I agree that *some* papers are good in ways that make them unpublishable by contemporary standards. But most of the advice in here is focused on making papers *better*, not just gaming the system!
In particular, I think these are good tips for any writer:
1. Use vivid examples, not boring abstractions.
2. Focus on your main idea, not irrelevant side-quests.
3. Be charitable instead of taking cheap shots.
The only reason why "Killing, Letting Die, and the Trolley Problem" might not get published today is that it doesn't cite much. (Only six things -- and Foot's paper is cited incorrectly.)
I'm not saying it's *good* that contemporary journals expect so many citations. But it would be malpractice to tell grad students to write like it was 1976. If they want research jobs, they *need* to get things published!
Still, I take the point that we don't want everyone writing for referees. I myself am kind of sick of that style of writing. That's why I'm on Substack! And it's why I'm spending more time writing books, where the author has a lot more freedom.
I wasn't thinking about gaming the system. I suppose I was denying that it's a way of making papers better—at least for many of my favourite papers. What's so nice about (many of) Judy's papers (along with being wonderfully clear) is that there isn't any signposting, there isn't any "here's my thesis that I will defend," there are digressions into side-points etc. (E.g. "Killing, ... " doesn't even have an introduction—not by contemporary standards). Instead, she takes your hand and takes you for a walk, pointing out some things along the way. And my point was that if Judy had followed this advice way-back-when, all her papers would have been stylistically different and much worse for it. (I use Judy as an example, as I know you like her writing.)
(PS. I wasn't thinking about citations, either. Though I do love how it has five (?) citations.)
Good point. I do love being "taken for a walk." My favorite Frankfurt papers have that quality, too.
In defense of my honor, though, I swear that I didn't recommend adding signposting! I think papers these days use way too much signposting -- usually as a way to make up for not having a good arc.
This is great advice. I wish I had read this in 2021. Here's something new I learned from this post that I've been thinking. I always thought the point of the introduction was to sew tension into the existing literature and then resolve it. The rest of the paper then just fills in the details. Upon reading this, I realize that a lot of my favorite papers don't do that. After introducing the tension, the author says just enough to give you the impression they are going to resolve it, and they give you some sense of what the ingredients of the resolution will be, but you don't really know how they're going to come together until later in the paper. It's hard to pull off in much the same way jazz is hard to pull off.
Love this comment.
You raise a more general point, which is that we sometimes don’t appreciate what our favorite papers are actually doing. When I started studying my favorite intros, I was *very* surprised to see how slow and restrained they were.
The first sentence of Jake Nebel’s “The Sum of Well-Being” isn’t about sums. It’s not about measure theory. It is literally just this:
“A person’s well-being is how good things are for that person.”
Here's a great paper I read recently by Tyler Brooke-Wilson,
https://philpapers.org/rec/BROHIP-2
I didn't notice when first reading it how little he tips his hand in the introduction. This post prompted me to reflect on it. He introduces the conventional wisdom on encapsulation, he tells you he's going to show that it doesn't explain the thing everyone thinks it explains, but he doesn't really tell you what the key move is going to be that undermines the conventional wisdom. Still, he gives you the impression he's issuing a credible threat.
Wow, that intro is killer. "Credible threat" indeed!
You mention that a 6000 vs 8000 word article is more likely to be accepted, ceteris paribus. But do you think that it’s a factor thats worth strategizing around? Should people aim to send 6000 words instead of 8000 to journals with an 8k suggested limit?
I had heard that going over the word limit incurs a penalty but this is the first time that I’ve heard that going under it gives you a boost. So I am not sure how to weigh it.
Of course you should eliminate unnecessary fluff etc from all your papers, but most papers seem to have some extra “bloat”, so my impression is that the “bloat” might even be considered a part of the academic style. So there could be a tradeoff between style and reducing word counts in terms of probability of acceptance.
I think short papers have better odds, partly because journals like brevity, and partly because short papers are more likely to get a quick decision.
If your paper is 6k words, reviewers are much more likely to accept the job, and they're also going to do it faster. I really can't overstate how important it is to get fast decisions -- it helps you get more feedback and send to more journals!
Also, in my experience, short papers are easier to do R&Rs for, since you have more room to add things.
This is excellent advice, and a real service to the profession. Thanks, Daniel!
One small point to add. I myself have followed the advice of 'publish on topics people you respect have been publishing on.' But philosophical temperaments differ, and I think some people are at their best forging out into (relatively) new territory. That can be a rough road, but it can also produce some beautiful and original work. It would be a shame to discourage overly much.
When forging into new territory, you have less guidance on what people are interested in from looking at the literature at what people have already been paying attention to. So for this sort of work, it is particularly valuable to present it as talks, talk over it with many people at the bar, etc., to get a sense of what are the things people find fascinating about the new topic and what are the ways you can motivate people to care about it even though they haven't thought about it already.
The advantage of publishing on topics people have already been publishing on is that you get a shortcut to getting an audience to care, because you can already see how existing papers on this topic get audiences to care. (And the existing papers may have even already created an established audience for another paper on the topic.)
Agreed on both counts, Kenny & Jake.
By the way, since you're both here, let mention a couple more papers with highly publishable intros:
Easwaran, "Decision Theory Without Representation Theorems"
https://philosophersannual.org/34articles/easwarandecision.pdf
Zuehl, "Kolodny Against Hierarchy"
https://philpapers.org/rec/ZUEKAH
I think these two intros make for a nice contrast. Kenny cuts right to the point, whereas Jake takes his time setting the scene. If you're a PhD student reading this, it might be worth pondering why these papers are (successfully!) framed so differently.
Thanks for this helpful article!
Have you ever brought back a paragraph from the graveyard?
Funny story.
Last year, my advisee was having trouble writing an intro, so I showed her the Grice & Barnett section of this guide. Then, to demonstrate the value of rewriting, I shared the painstakingly rewritten intro of my latest paper along with a version from the graveyard.
To my horror, she told me that the new version was "Grice," and the old one was more "Barnett." And she was right!
I definitely had to do some exhuming after that.
One more thing! Sometimes I find myself graverobbing old papers for material to start new ones. The seed of "Caring Who" was a page cut from "Each Counts for One."
What about you?
I don't have a graveyard - when I delete stuff it goes into the void forever. But I have deleted entire sections after realizing I can't flesh them out as much as they deserve within the paper's constraints, and resurrected them as full-fledged pieces of their own. In one case, I deleted two sections from the same paper and used them both in a new one.
What you say is probably quite right, but I rather hope people don’t follow it. And I doubly hope reviewers don’t take it on board. And that’s because many of my favourite papers would do terribly by these lights, e.g. “Killing, Letting Die and the Trolley Problem.” (How sad it would have been if Judy had received and followed this advice way-back-when.)
I agree that *some* papers are good in ways that make them unpublishable by contemporary standards. But most of the advice in here is focused on making papers *better*, not just gaming the system!
In particular, I think these are good tips for any writer:
1. Use vivid examples, not boring abstractions.
2. Focus on your main idea, not irrelevant side-quests.
3. Be charitable instead of taking cheap shots.
The only reason why "Killing, Letting Die, and the Trolley Problem" might not get published today is that it doesn't cite much. (Only six things -- and Foot's paper is cited incorrectly.)
I'm not saying it's *good* that contemporary journals expect so many citations. But it would be malpractice to tell grad students to write like it was 1976. If they want research jobs, they *need* to get things published!
Still, I take the point that we don't want everyone writing for referees. I myself am kind of sick of that style of writing. That's why I'm on Substack! And it's why I'm spending more time writing books, where the author has a lot more freedom.
I wasn't thinking about gaming the system. I suppose I was denying that it's a way of making papers better—at least for many of my favourite papers. What's so nice about (many of) Judy's papers (along with being wonderfully clear) is that there isn't any signposting, there isn't any "here's my thesis that I will defend," there are digressions into side-points etc. (E.g. "Killing, ... " doesn't even have an introduction—not by contemporary standards). Instead, she takes your hand and takes you for a walk, pointing out some things along the way. And my point was that if Judy had followed this advice way-back-when, all her papers would have been stylistically different and much worse for it. (I use Judy as an example, as I know you like her writing.)
(PS. I wasn't thinking about citations, either. Though I do love how it has five (?) citations.)
Good point. I do love being "taken for a walk." My favorite Frankfurt papers have that quality, too.
In defense of my honor, though, I swear that I didn't recommend adding signposting! I think papers these days use way too much signposting -- usually as a way to make up for not having a good arc.
I rescind all reference to "signposting."