NOT SO ABSTRACT ANY LONGER

I don’t have all that much to say about abstracts and titles that I haven’t already said in other portions of this guide, to be honest! That's out of character for me, I know. Truthfully, I find writing abstracts powerfully dull, even in the best of times.

That’s a shame—abstracts are incredibly important. Not because they contain groundbreaking content (in fact, nothing in an abstract should be new or said substantially better than it is in the paper), but because abstracts are primarily what search engines still use to decide which papers to present to a prospective reader. Plus, they’re a large part of how prospective readers decide whether to read a paper at all.

So, yes, abstracts have an outsized influence on your paper’s impact—whether they deserve it or not. The romantic in me resents how calculated I have to be with this part of a paper, but the realist in me knows it matters, and that tends to make me drag my feet.

Another reason I tend to loathe writing abstracts is they usually have to be so bloody short—most journals impose a tight word limit on them, and they may also require they contain very specific information (like an explicit “so what?” section). That means abstracts, like Introductions, have to be dense.

So, more so than even for Introductions, I’ve had to really abstract (pun intended) the process of writing abstracts to make the process tolerable. Here’s how I approach it:

1. First off, I don’t even bother writing an Abstract draft until the entire rest of the paper is in its second draft or further. Yep, feel free to procrastinate on this until then! Dr. B says it's ok. Abstracts are supposed to be high-level summaries of the entire paper—the paper needs to be pretty far along for writing such a summary to make any sense.

2. Second, I aim for the “heart” of my abstract to be just 8 sentences—I plan to pull exactly two sentences of content (synthesized and phrased originally!) from each of the other four sections (Introduction, Methods, Results, and Discussion).  

3. For each pair, I ask: What is the “bumper sticker” version of this whole section? What would I want a reader to walk away with, even if they never read the paper itself? I also try to make sure I’m “advertising honestly”—telling my audience exactly what they’ll get, in greater detail, if they do decide to read on.

4. Concurrently, I brainstorm keywords—terms or phrases someone might naturally search for if they want papers like mine. Then, I try to sneak as many of these into the abstract as I can, because, yes, this is search engine optimization. Welcome to modern science; this (still) dictates how most people find research papers. It’s better to play the game well then to not play it at all. Besides, you usually have to provide a list of keywords anyway, so this is “double-dipping” on that work.

5. If a section only needs one sentence to be fully incapsulated, great. If another needs three, I force myself to compress elsewhere to stick to eight sentences, if I can. A shorter, tighter abstract is always better.

6. Don’t try to include all your Results in your abstract. Instead, pick your two-ish most impactful ones to highlight. I know that probably hurts. Think of them merely as “teasers;” they promise even more of the same to someone who reads your paper.

7. Pet peeve alert: Please don’t reuse sentences verbatim from your paper in your abstract! It’s obvious. It’s boring. It suggests to me, at least, that you’re not quite as passionate about your work as you probably are. There are always at least two good ways to say something!

Inevitably, eight sentences is never quite enough to make the result feel complete. However, by the time I have my eight sentences, I usually have a clear sense of which parts are feeling "thin." Often, for me, it's the Introduction (perhaps I need a third sentence to bring in the "Large Known" or "Justification.") or the Results (there's a third-tier outcome I want to highlight). I may also end the abstract, much as I like to end my Discussions, on a "next step," which may require a third sentence as well.

In summation, I think of abstracts as the “elevator pitches” of science, an “audition” you have to pass for others to keep reading. Auditions are rarely fun, but I do appreciate they are necessary, so we better come ready to them. Punchy, terse, and unapologetically pushy has always felt like the right way to tackle abstracts for me.

FROM MERELY A "WORKING TITLE" TO A TITLE THAT WORKS

We close the main portion of this guide at the beginning, ironically. I told you I was dedicated to my "working backwards" system!

I have three main pieces of advice for Titles, but, fair warning: they’re opinionated (ok, ok, more opiniated than usual!). There're those who would disagree with me (my Ph.D. advisor sure did!), and that’s the way it should be. There’s no one right way to do science. But, hey, you’re here, so I presume you’re open to my opinions, so let’s do this.

First, though, let’s rehash how someone will most likely decide to read your paper:

1. They’ll likely start with a topic of interest.

2. They’ll then (probably clumsily) condense that interest down into a keyword or three.

3. They’ll dump those into a search engine like a gambler pulling a slot machine arm and hope for the best.

4. The search engine will spit back a slew of article titles, which the person will now skim as fast as humanly possible. Time is $$$ for all of us!

5. As they do, they’re probably asking themselves two questions for each title they see: 

   1. Does this sound relevant? and 

   2. Does this sound minimally tolerable to read?

6. If (and probably only if) a title earns two yesses those questions, the abstract will probably get the same treatment. Which is why, even though I hate writing them, I still put extra care into my abstracts. 

7. Then, often, your figures and tables will be the next consideration (I told you not to sleep on those!).

So, if you’re playing along at home, that’s:

Interest area ➡️ Keywords ➡️  Title ➡️  Abstract ➡️  Figures/Tables ➡️  Rest of your paper

In other words, to get read cover to cover, your paper will often have to first successfully cross five “security checkpoints." The more you can do to make your paper fit into more interest areas, come up in more searches, and pass more of these “two-question tests,” the better. Sadly, given the oceans of published research, many papers will fail to cross all these “checkpoints” for most prospective readers. We all gotta filter to survive!

This means we need to seize every opportunity to teach people something, even if it’s less than we wanted to. That starts with the title, the first thing most readers will encounter (and, unfortunately, potentially the last too).

In my view, there are two common categories of title out there, one that often teaches the reader something, and another that often doesn’t. Guess which I prefer??

The first, which I call “methods-oriented titles,” center what you did. Here’s an example: “Determining the role of nitrogen limitation in seed dormancy breakage in black raspberry.” This title contains some keywords, and it definitely tells us what we can expect from your paper...

...However, it totally places its focus—wrongly, I’d argue!—on what the researchers did (or explored). Unless your methodology was novel or imaginative, and those novel elements can make it into your title, such a title likely won’t teach me anything.

Contrast that title with what I call a “results-oriented title.” Such a title centers what you found. The same study might instead have had a title like “Nitrogen-limited black raspberry seeds break dormancy slower, corroborating the N-loss hypothesis.” This title does no worse in terms of including keywords and setting expectations, but it teaches us something too—black raspberry seeds need available nitrogen to germinate! We know that now! If I stopped reading here, I almost can’t help but walk away with that tidbit, whether I wanted to or not. It’s “force-fed learning" of the exact kind required by the blistering pace of modern science.

So, if my first piece of title advice is “stuff in a few keywords so the algorithm blesses you,” my second piece would be “at least tell us something you learned.” We’ll appreciate you for your efficiency, if nothing else.

The above advice ensures your paper gets found and has at least some impact. My last bit of advice relates to getting your paper read. I said earlier that a title may dictate whether a reader concludes your paper is going to be a chore to read. It’s our first glimpse at your writing; can we penetrate it? Would we even want to try? If you don’t silently judge the readability of papers using their titles and abstracts and act accordingly, you’re a better person than me! But I doubt I’m alone.

So often, it feels as though titles are among the driest sentences us scientists produce. And, look, I get why. Science is supposed to be serious, dispassionate, professional, to-the-point, et cetera. A “by the book” title fits squarely within that paradigm.

…But maybe it’s just who I am; I’m incapable of taking myself and my work so seriously that there is no room, even in a title, for a little joy or passion. So, that’s my third piece of advice, which you may absolutely take or leave: Try to find a way to add a little “sparkle” to your titles.

For example, here’s the real title from one of my papers: “Bearing fruit: flower removal reveals the trade-offs associated with high reproductive effort for lowbush blueberry.” It’s a pun—plants bear fruits (produce them), but they also bear them (cope with their costs), which relates directly to the goals of my study.

Now, is this a million-dollar pun? No. But if it made even one person smile or, better yet, read my abstract or paper when they otherwise wouldn’t have, that's a win. It was my way to “virtue-signal” that I intended to teach and engage with my paper. I firmly believe science can and should try to do both.

I have a little more to say about titles, but it isn’t as punchy nor insightful, so let’s do it rapid-fire-style:

• Good titles tend to be short. This is because most journals require it, sure, but also because snappy titles are simple titles, and simple titles are good titles. Here’s the one place in science writing where I actually appreciate word limits! Here’s my challenge to you: Stick to 16 words or less, even when not forced to. If you can succeed in fewer, even better.

• There're some words you should banish from your science vocabulary, at least when it comes to titles. These include (but are not limited to) “impact,” “effect,” “explore,” “examine,” “understand,” “factor,” “change,” and “different.” These words are “crutch words”—they are the imprecise words we scientists too often reach for when being more precise feels like too much work in the moment. If something had an “impact” on something, why not just say what that impact was? Compare “Calcium’s impact on joint health” to “Calcium spurs more durable joints.” The latter is clearly better, right?

• Just because the title is often the last thing we write doesn’t mean it should be an afterthought. I’d argue a whole project team meeting could and should center around just workshopping the title.

• If you have a target journal in mind, scan the titles of papers from the most recent issue. You might get title ideas from papers that have successfully gotten published. There’s no shame in learning from what works.

Your paper’s title is the first impression you’ll leave on the scientific community, so it pays to make it sharp, honest, and—just maybe—a little inviting. After all you've done up to this point, your work deserves to be read. Make sure your title doesn't trip you a the first hurdle!