Deconstructing Gawande – why narrative and structure are important

By Ed Yong | August 23, 2010 9:00 am


According to all sorts of “how-to” sites, one of the most important rules for online writing is to keep it short. Well, try telling that to Atul Gawande and his legion of readers. Known for his superb long-form essays and books, Gawande recently published a masterful piece in the New Yorker about death and dying. I plugged the full text into Word and it weighs in at a monstrous 12,000 words. This is a feature that makes Wired articles look like paragraphs.

And yet, its copious length flies by. I read the piece in 20 minutes, with no distractions except for the occasional need to reach for some tissues. I’m not alone in thinking this; the piece was passed all over Twitter, with several people noting how readable it is despite the excessive word count. And bear in mind that all of these people read the piece on screen (and I digested it as a single, scrolling column). How does it succeed?

Part of the answer doubtlessly involves the subject and Gawande’s own experience. Death is a topic that is close to our hearts, either through experience or contemplation. There can be few issues that are more guaranteed to stir an emotional reaction in a reader. Gawande also has an inbuilt edge in that he is a doctor first and a writer second. He absolutely knows what he’s talking about and he has years of stories and medical experience to draw from.

But I would argue that even these massive advantages do not guarantee a good read. In a similar position, many writers would still struggle to keep a reader’s attention rapt for a 12,000-word magazine article. To me, the secret of Gawande’s piece is its masterful structure. This is a textbook example of how to use narrative to sustain interest, regardless of length.

Let’s break it down.

The piece is about death and how people deal with dying. It’s an expansive topic and while Gawande covers a large amount of territory, the piece really only has four (maybe even just three) main ideas:

  • Caring for terminally ill patients is a big, neglected and increasingly important issue.
  • Hospices are about much more than just letting people die.
  • Discussions about the end of life are important for terminally ill patients and their loved ones.
  • Doctors need to have these discussions with their patients too.

And that’s it. Four ideas, and the last two could realistically be grouped together. You could write all of that on a PowerPoint slide.

This is not to downplay the quality of the article. The piece is readable exactly because it hinges around a small elite set of messages. Each one has stories, statistics and facts that support it, but these are side dishes that are used to flavour a meaty, central ingredient. Cooks risk spoiling a dish if they try to pack in too many textures and flavours; likewise, writers can make their articles unpalatable by pursuing too many tangents or packing in too many ideas.

But simplicity is not incompatible with depth. Each of Gawande’s four main concepts could be the subject of a feature article in themselves and, in fact, they are. I’ve tallied up the word counts spent on each one and they’re all the length of a 2-4 page piece in a normal magazine (respectively, they come in at 1700, 2300, 1300 and 2500 words).

These four sections are all obviously united by a common theme. But to hang together in a single feature, they need more than that. Gawande achieves this by using the tale of a terminally ill cancer patient, Sara Monopoli, to frame the four topics. It is obvious enough to use real-life stories to illustrate the theme of death and Gawande’s experience gives him plenty to draw from. But his critical move was to use a single story to frame all of the others.

We meet Sara three times from her initial diagnosis through a gruelling series of choices and treatments. Each episode provides anchors for the rest of the piece, introducing the key themes while providing an emotional centre. In the brief introduction, we see the events that lead up to a terminal diagnosis and we see this as a problem that affects real people. From there, Gawande segues veers off into his first two topics – the importance of end-of-life care and the role of hospices.

We cut back to Sara for a hefty segment in the middle of the piece (and if you haven’t read it, you should probably do so now). We learn that her treatments aren’t working and we hear about the impossibly tough decisions that she and her family have to make. Again, this introduces the importance of discussions between patient and loved ones, and between patient and doctor – the third and fourth core themes of Gawande’s piece. Finally, the ending (of Sara’s story and the feature) is both fitting and inevitable.

The whole piece works roughly like this, with word counts attached:

  • Sara’s story – an introduction (1,000), which leads into…
  • The first act – the importance of end-of-life care (1,700) and the role of hospices (2,300)
  • Sara’s story – an interlude (2,600), which dovetails nicely into…
  • The second act – the importance of discussions for loved ones (1,300) and doctors (2,500)
  • Sara’s story – an epilogue (600)

Of course, there’s more to it than this. There’s the skilful storytelling, the deft use of anecdotes and case studies besides Sara’s, the sense of underlying academic rigour, and so on. But all of that hinges upon this brutally efficient structure – without it, the other elements would collapse like a body without a skeleton.

I’ve always said that to write well yourself, you need to constantly read good writing. So read this.

Photo by Vees


Twitter.jpg Facebook.jpg Feed.jpg Book.jpg


Comments (22)

  1. Ed, thank you for the link. I also read it with some tears. Your analysis is apt. But I think it leaves out one thing: the author’s honesty and the vulnerability he allows to shine through it. The subject matter is important and emotional, the structure efficient, the annecdotes give it a human face, the writing is competent. All of that is there. But what is also there is the power of truth and a willingness to look deeply to find it.

  2. your analysis is spot on. but, if everyone took the same advice the landscape would be homogeneous and there’d be opportunity for those who filled empty niches.

  3. Yes, this was Gawande at his best, and very reminiscent, for me at least, of his masterful story on the Apgar score a few years ago.
    (Some pieces just stick with me and I still can’t seem to get that one out of my head.)
    Lilian also makes a good point. Gawande risks a lot by putting himself in the story, and indeed he’s included himself in some of his less successful articles in unfortunate ways. His article on itch, for example ( fell apart with his intrusions at the end. They seemed almost hokey, a shame after one of the most evocative intros I’ve ever read.
    I think he uses himself best when he exposes his own insecurities and self percieved shortcomings. He becomes no longer a doctor talking down to an audience, but a participant in the larger story. And as such, he’s at the mercy of the same factors as everyone else, patient, family member, doctor, you name it.
    Here’s another nice first person take on end-of-life care — more centred on the family member issues and emotions (from The Atlantic).

  4. Joe

    It’s a catch-22 that you have to break down great writing to figure out why it is so effective and beautiful, but when you do the magic of the experience seems tarnished.

    Few things I have read this year were this moving and utterly important to think about. I am glad to see it is maintaining legs. Should be required life-reading.

  5. I’m looking forward to reading it in its entirety, but what Lilian said plus the fact that it’s written in highly approachable language. There aren’t a lot of acronyms used, or terms that need ‘translation’ in this piece. I was going to say that when writing a long blog post, many people forget what’s standard in print journalism, namely subheads. Of course Gawande doesn’t use them. BUT – the raised caps that occur every four to six paragraphs perform the same function, making the piece less intimidating because you know you can read it a section at a time if you have to and still find your place if you’ve been interrupted by a phone call or email (or, you know, someone asking you something in PERSON, if anyone still has the temerity to do that). 😉

  6. Say what you will, but I think Gawande’s article is overlong. Wonderfully written, reflection inspiring, touching – but overlong

  7. Great comments folks. And for the record, I’m not arguing that structure is the only element to the success of this piece – it’s just the one I wanted to unpick in this particular case.

    Razib – agreed. This wasn’t intended as a template for others to fit their long-form pieces to. Other pieces will demand other structures. The point was how structure makes or breaks a story.

    And Joe, I disagree that analysing a piece in this way diminishes the magic for me; if anything, it helps me appreciate what Gawande did more deeply. I only really wrote this (after all, who the hell am I to dissect the work of a better writer) because I noticed something about the structure and thought, “Cool.” In the same way, understanding how evolution and aerodynamics work in no way diminishes the sight of a bird in the air…

  8. I agree with everything Ed says, and also with the point that Gawande is showing an unusual amount of vulnerability. Compare this with his McAllen piece, which is pure, calculated reason. It’s all head. This one has heart.

    For me, the most remarkable thing about this piece was the way Gawande caused me as a reader to experience the same emotional process that Sara and her family did. I knew this was a piece about dying, and I knew Sara was going to die just as surely as Sara and her family did. But in spite of every rational and obvious and inevitable thing I knew, I kept hoping throughout the article that she would live, almost until there was no more article to read. (“Almost,” because I stopped hoping right around the time her family did: when her mother stood over her protectively and told the doctors to leave her the hell alone.) This was the same ludicrous hope that motivated all the characters trying to keep her alive at any cost. Gawande didn’t just write a piece about people irrationally refusing to accept a person’s death. He made his readers (or at least this one) refuse to accept it, too, and then reluctantly give up and let go.

    I’ve been trying to figure out how he did this to me, and best I can tell it’s through the same tricks that great novelists use. Examples:

    * The entire piece is contained in the first sentence: “Sara Thomas Monopoli was pregnant with her first child when her doctors learned that she was going to die.” Gawande immediately introduces us to a young, appealing, vulnerable main character. We feel for her, maybe even identify with her. Then, in the next beat, he tells us the plot: She is going to die. As people with feelings, we don’t like this, and as readers, we know that plot demands conflict. The primary conflict, we can tell from this first sentence, is going to be the fight against death. But since this is a conflict that people can’t win, Gawande’s article will need a more satisfying (i.e., winnable) secondary conflict, which it does have: *how* best to fight death. The family ultimately wins this secondary conflict. But it loses the primary one, as it must. This is powerful, universal, ur-narrative stuff, the kind of structure you see in Norse mythology, where the hero must die but chooses to do so “victoriously” (and this is what makes him a hero).

    * Foreshadowing. There are threads throughout the piece that Gawande presents initially as simple, interesting points, then picks back up at moments of extreme emotional intensity, giving them new significance. For instance: “As for last words, they hardly seem to exist anymore. Technology sustains our organs until we are well past the point of awareness and coherence.” Then, many paragraphs later: “‘Neither can I,’ Sara murmured. Only later did the family realize that those were the last words they would ever hear from her.” Or, describing the death of a different patient, Lee Cox: “The emergency medical technicians and firefighters and police rushed in. They pulled off her clothes and pumped her chest, put a tube in her airway and forced oxygen into her lungs, and tried to see if they could shock her heart back. But such efforts rarely succeed with terminal patients, and they did not succeed with her.” And later, as Sara (and the article) is nearing the end: “A few minutes later, firemen swarmed up the stairs to her bedroom, sirens wailing outside.” The images mirror each other; the medical professionals “rush” and “swarm” in. Subconsciously, we get the signal that their efforts will be as fruitless in Sara’s case as they were in Lee’s.

    The reason I bring up the McAllen piece as a comparison is because this piece could have easily have been written the same way: Gawande could have made a cool, well-reasoned argument about why hospice care is often preferable to what some doctors call a “flogging” at the end of life, and about how medical policy ought to reflect that. In other words, he could have basically done a long-form version of the recent palliative-care paper in the NEJM. Instead, he decided to engage readers’ emotions, and for what my opinion is worth, I think in the process he wrote a literary masterpiece. Anyone who still thinks of him as a doctor-who-writes will now have to admit he is very much a writer. I’d be interested to know if other people experienced this piece the same way. And I’d love to know if Gawande was consciously thinking about this kind of stuff as he wrote, or whether he just has an intuitive feel for how to move people’s emotions.

  9. Ed, this is great analysis from a narrative and conceptual standpoint. I wholeheartedly agree that an understanding of the power of narrative is absolutely essential to good science writing.

    Gawande excels at beginning his article with an emotionally powerful anecdote and keeps the story moving along by introducing incidental details. One stunning aspect of his work: the reader could stop at virtually any time and still come away understanding his primary message. Yet, his writing never dead-ends or falters because he allows it to gently meander by following what seems to be a normal, associative but still logical train of thought. We all think of one thing and then are reminded of yet another—Gawande uses this digressive technique to great advantage.

    In addition to his writing prowess, I would like to suggest that there are other equally vital characteristics of this piece that make it eminently readable as well. For one, let’s remember that visual presentation of material is crucial to ease of reading, comprehension and retention. The success of browser add-ons like Readability have repeatedly reminded me that *how* information is presented is often just as important as what is presented. (Speaking as a designer, this is a mantra that I wish more bloggers, web designers, and online publishers adhered to…)

    Along these lines, Gawande is well served by the simple one column format, as much as he is by font size and choice of serif typeface. That his article eventually appears as only text on the screen with no additional visual distractions, after the reader scrolls past The New Yorker information and its associated ads, further enhances the piece. This clean, simple design focuses the reader and removes mitigating distractions—distractions, I daresay, that stop many a web reader practically before they begin. Indeed, if an online publisher wants to have their work read, then simplicity and ease of reading combined with compelling content are a winning combination.

  10. Peter Demain

    Mary: TDLR. ( Just kidding.)

    Some time ago an editor I got commissioned to write an article on mental health stigma. You could write a book on that matter, but I was limited to 1000 words as apparantly ‘people get bored and won’t read more than that’ . That seemed rather like assumptive generalizing back then, but it’s an attitude that pervades journalism in general.

    If you have a good book (for instance Peter Cook’s biography) you end up reading thousands of words without realizing. Same with good fiction. That’s because the writing is excellent; it strikes a chord with you by virtue of informativeness, humour, or by enlivening a subject in such a way that you can understand and be entertained to the point of voraciousness.

    As diseased as journalism is at the moment (supplements, usual suspects like the Mail etc) there are still a minority around who can operate a camera and make it a passion to do quality work. Most today sit in offices all day doing copypasta from the press wires, or hashing together opinion pieces with spurious quotes from anonymous ‘analysts’. Whilst the salaried lot are on the whole better paid, they lose soul in the drudge fast. The talent they began with withers as a plant would upon lacking sunlight or water.

    Point is is that great writing can keep a reader’s interest. Now Ed I’m sure you’re different than the audience for my blog – I try to put science in now and then but it’s mostly satire/pop culture etc – but the same principles apply. If something is scathing, compelling or hard-hitting it gets read in full…but all the same (might be lack of confidence) the longest article present on my website is presently 1500 words. The average hovers around 900 – 1000 – I think based on this thread I might start doing longer pieces.

    In the maddening hysteria of day-to-day life, I believe people do yearn for and enjoy a meditative or fun time reading in peace. I don’t mean a paper, I mean (as Mary infers) prose so good that it approaches the artistic in terms of how it’s put together. Stuff with the author’s passion and soul behind it, written in a lively, interesting and fun way.

    We humans were not built to work as much as we do now. That’s why you get stress, breakdowns, depression – in 1932 Bertrand Russell wrote ‘In Praise of Idleness’ – 78 years on we’re reaping the results of a lack of idleness. I learnt enough about office journalism to never enter that trap; real journalism is going out, talking to people, doing photography, and being a responsible part of the community. That’s the idea anyhow. Sitting on your backside at a desk cannibilizing stuff everyone else reports, or regurgitating PR is not what the trade is meant to be.

    Imagine if journalism were like mining. The news is a massive, unending bedrock crammed with lode after lode of material. The corporations did to our ‘mining’ what Thatcher did to the unions in the 80s. It’s not about employment or good work now, it’s about profit and squeezing as much as possible from as few staff as possible in as short a time as possible. So the mines are mostly derelict, and most people in the ‘profession’ sit about and examine other peoples hauls, or the Fool’s Gold of PR and fabrication.

  11. I know, I know. Hey, I never said I shared Gawande’s skill….

  12. gillt

    My science writing instructor often used Gawande’s essays in class.

    Not that you’re implying such, but I’m not on board with the assumption that science/medical articles need to have a personal narrative component, which is what’s often pushed for in nonfiction writing programs…as if readers require drama to stay engaged. This is especially the case when you’re writing a shorter piece. Then it often seems forced and deprives the readers of good ‘0l science.

  13. Regarding human/personal narrative, I’m firmly in the “Apply when appropriate” camp. In this case, it would be sheer madness to not do it. In other cases, the science can speak for itself without talking about what the scientist’s dress sense is like and what their favourite food is.

    People often talk about personal narratives as if they suddenly make writing easy and turn the “interesting” dial up to 11. But I think it often has the opposite effect, because you need a lot of skill to be able to weave a human story into a scientific one without the reader noticing the seams. A notable case study where this was done right was Rebecca Skloot’s book, The Immortal Life of Henrietta Lacks (which, in hindsight, was the first of these deconstructions that I did).

  14. I work at our library semi-annual used book sale. Based on that experience I can confidently say there are THOUSANDS of tomes full of absolute rubbish published each year. Very few make me think. So Gawande’s article was exceptional.

    I have to say thank you to Ed — I would probably never have read the article had you not linked it here. I respect your ability to choose the good stuff out there and link to it. Thanks.

  15. gillt

    Funny you mention Rebecca Skloot. I went to a talk she gave to a group of scientists where she said the personal angle was the key to the whole story. She had done the research and now she needed something for people to relate to. Maybe that’s what people expect out of books and NYer length features. Maybe that’s what publishers and editors want. It’s hard to tell the difference.

    Sure the personal narrative is difficult which is why it’s often given center stage, often to the detriment of discussing science. For what it’s worth, I consider your posts a reprieve from all that. Do you write longer pieces?

  16. Sean Patterson

    Yes, narrative and structure are important but attaining deep self-assurance that life has no beginning and no end before we reach the end of our ‘skein’ should trump that. It never seems to, however, and I’m always amazed by that. So I guess I won’t start a blog on the real point…

  17. So on this question of human stories, here’s a very interesting case study. In this year’s ABSW Science Writer awards, which I helped to judge, two pieces jointly won the prize for Best Feature. Both were on cystic fibrosis: one by Robin McKie in the Observer and one by Helen Pearson in Nature.

    Two pieces, same topic, almost completely different approach. Pearson’s piece is led by the science, McKie’s by the human element. We honestly couldn’t decide which was better, despite a LOT of discussion. This is a great example of how different approaches to the same topic can work, as long as the writer is good enough!

  18. One reason this piece works so well is that Gawande takes us on his own journey.

    In the beginning he says, “Like many people, I had believed that hospice care hastens death…” but by the end of the piece his thinking about hospice and end-of-life care has completely evolved.

    This transformation–the surgeon changes his mind about hospice–gives the piece added force and authority.

  19. gillt

    Good reads! Most striking in McKie’s piece is the lack of Collins’s name. Pearson’s piece at times felt like reading a journal review of Cystic Fibrosis, which biased me in his favor. You’re right, they are great case studies of what we’re talking about.

    Another nice example of combining a sense of the dramatic and personal while keeping the science front-and-center is Richard Preston’s “Demon in the Freezer.” He’s the “Hot Zone” author.

  20. Good stuff, thanks Ed.

    One day when I’m big I want to write as well as Gawande….

  21. Brian Too

    Perhaps this is what science communication really needs. Let’s just promote superior writers like Gawande, or superior works like this article.

  22. What a great article and comments! Together they gave me lots of things to think about for a project I’m currently working on looking at narrative structure.


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Not Exactly Rocket Science

Dive into the awe-inspiring, beautiful and quirky world of science news with award-winning writer Ed Yong. No previous experience required.

See More

Collapse bottom bar