I’m interested in taste, as in: what we like, what we don’t like, and why. Occasionally I read books on this topic, such as You May Also Like: Taste in the Age of Endless Choice by Tom Vanderbilt.
This piece of light reading occupied most of my summer vacation last year, and as I read I made some notes with the intention of writing what I usually write prompted by books: a review that turns into me pontificating on tangents.
In this case I thought of the, to me, frustrating way the book was structured. If you’ve ever seen one of those easily digestible science documentaries on the secrets of sugar, alcohol, the Aztecs, the moon, the history of mathematics or How Your Coffee Tricks You Into Watching Netflix you know what I mean: affable host presents topic, they visit some researchers and talk about their work, do some kind of simple experiment and the host tries something themselves and talks into the camera about what they’ve learnt. If it’s about anything health related you can be sure to see a shot or two from right before bed.
It makes sense that TV is like this. It’s typically only an hour at most and must be entertaining, easy to follow and contain visual variety (i.e. stuff happening on screen). Of course the TV medium avoids abstraction, complexity and theory in favor of easy conclusions and amusing tidbits.
It makes less sense to me when a book follows the same pattern, which is what Vanderbilt’s book did. It was full of nice little stories (about the author talking to researchers etc.) and details but avoided explaining anything in depth.
Just an indicative example: there was this section when he visits Netflix’s office and comments that they use Latent Dirichlet Allocation in their recommendation calculations. Now, it’s possible to give some basic explanation of what that is in a few paragraphs (I tried to explain the basic idea behind principal component analysis here and I think it works) but Vanderbilt doesn’t even try to do anything like that. It’s just there to say “look how technical this is!” with a chuckle. No indication that a reader might be capable of understanding the first thing about it. It might as well be magic.
While not important in itself, that was nonetheless symptomatic of what I saw as a disappointing lack of ambition for a book supposedly about “Taste in The Age of Endless Choice”. I had thoughts of an article called You May Also Like Meatier Books, setting up a typology of pop-science books based on varying breadth and depth. But in the end I didn’t write it because I get way more ideas for posts than I can manage to finish, or even start.
So why am I writing this now? Because I picked up another book on taste and liking for this year’s light summer read: Hit Makers by Derek Thompsom.
This time I was more aware of the pattern beforehand and noticed this book conforming to it to such an extent that it became outright funny.
A review in which man yells at book for not being what he wanted
The blurb on Goodreads calls Hit Makers “a groundbreaking investigation” into the nature of hits and why things become popular. That’s a big promise. I hoped Thompson would deliver on it and produce an ambitious yet accessible work on taste with some ideas substantial enough to chew on, or at least a nice synthesis of existing thought on the matter.
I’m sad to say that he did not. The book isn’t a filling steak dinner, it’s a bowl of jellybeans. Thompson provides story after story after story about hits of various sorts: Brahms’s Lullaby, Barack Obama’s “Yes, we can” slogan, Star Wars, Rock Around the Clock, the impressionists, Fifty Shades of Grey, Justin Bieber etc. They’re all interesting and highly readable but there’s hardly anything else there. As one Goodreads review puts it:
This is not the dry academic book, it uses very good storytelling and lots of techniques to maintain the reader interested. On the other side it encompasses too many perspectives, approaches and cultural production that makes it impossible to deliver the promise made by the title. There’s really no science in these pages, but only lots of stories, causalities and correlations.
Thompson’s got Gladwell’s Disease. He purports to write somewhat exhaustively about a topic but the topic itself seems little but an excuse for telling stories.
Yes, you need examples to present your ideas in a convincing, user-friendly way. Pure abstraction isn’t just hard to follow and relate to, it’s often dreadfully boring. Where Thompson goes wrong, in my opinion, is when he forgets the purpose of examples. They’re supposed illustrate concepts, make logical points land, and bring abstractions to life. They’re means to an end. They’re there to serve the ideas, and not the other way around. In this book they’ve taken over the whole show. They’re too many and too long — much longer and more detailed than they need to be to make the point. The theoretical complexity it would take this much storytelling to bring down to comprehensibility just isn’t there.
Thompson goes on for dozes of pages before moving on from “people like a balance of the familiar and the new” as his central insight. Yeah, I get it. I knew that before and even if I hadn’t it’s a conclusion complex enough for a few paragraphs. It’s also intuitive enough to not need reiterating from sixteen perspectives. Get the duck on with it, please. And that’s the typical pattern: we get some stories and a pat, often trivial conclusion that didn’t need all those examples and minutiae to be convincing or understandable. The only explanation I can find is that he’s not actually telling me stories because he wants me to understand something but because they’re interesting stories. His attraction to personal stories and mini-biographies is strong enough to sometimes border on the parodic. For most of the book it feels like every goddamn chapter opens with some influential/interesting person’s life/success story in about one thousand words. His love of storytelling is so strong that some neat stuff he apparently couldn’t fit into the main narrative are included as interludes, including the history of teenagerdom, his own love affair with acting, and visits to the Strand bookstore in New York. They’re clearly little labors of love. Or, if you want to be mean, a flock of unkilled darlings.
The pattern is characteristic of this whole genre of hyperpalatable nonfiction. It can be described as an inversion of the relationship between means and ends: the supposed end of coming to a conclusion about some topic instead becomes the means to write a lot of very competent journalistic prose around individual examples that ought to be the means of explicating difficult ideas but turn into ends in themselves.
A book on a topic requiring some theoretical exploration — and where the writer hasn’t fallen in love with their storytelling to the extent that they no longer know its proper place — is structured like one big, complex thought whose content gradually emerges and solidifies as the reader makes their way through. Concrete examples and illustrative stories are strategic tools placed at critical junctures to bring points home, either because the reader needs an idea to be explained, because they need to be convinced of its validity, or just to make it “feel real”. They are typically not longer or more detailed than they need to be to perform those functions.
Think of it like this:
What happens when someone replicates this structure but focuses on the wrong features? What if you miss the macro-structure and instead just love to write the little illustrative stories? The stories grow in size and number and they stop being subservient to the larger structure, which starts to buckle and eventually disintegrate under the weight. In the end the structured thought the book supposedly expresses has been reduced to a vague “theme” or “topic” the writer meanders about in — a strand of thematic commonality just barely holding a bag of intellectual empty calories together.
This isn’t a book with a thesis that uses stories and tidbits as delivery methods. It’s a book of stories and tidbits that uses the expectation of a thesis as packaging.
The exception that proves the rule
The chapter called The Viral Myth is interesting because it breaks the pattern a little bit and adresses a specific question directly. In it, Thompson argues that “things going viral” is a myth and nothing really gets popular the way the “viral” metaphor suggests. His point is well made: things really don’t become hits by spreading from one person to a few, and from those few to a few more each etc., like a real virus does. Instead so-called “viral hits” get popular because they on at least one occasion get transmitted from one to many in what looks very much like a traditional broadcast.
This is a good point even though I think he’s taking the word “viral” a little too literally. It doesn’t necessarily describe spreading only by one-to-few transmission. In my mind it means being propelled primarily by other sources than its initial sender, and not depend on a high-reach, high investment initial broadcast for momentum.
In this chapter there’s a seed of a much more theory-heavy, even mathematical approach to the same issues, one that I would have absolutely loved to read. Unfortunately it’s the exception that proves the rule (and I’m actually using that expression correctly here). His mathematical thinking skills seem shaky. This botched graph on page 188 is an example:
It’s a depiction of what “virality” would look like, except that what he was going for, judging by the description in the text, was this:
It might not look that different, but mathematically they describe very different structures and not noticing that is a sign of, well, shakiness. To be blunt, Hit Makers isn’t a book for the mathematically literate. At one point it mentions power law distributions which can be easily explained by a graph and a few sentences but instead Thompson opts for saying that it’s like a lottery where most win nothing and a few win hugely, which is such a low resolution version of the truth that it’s highly misleading.
After the virality chapter the anecdote-after-anecdote style comes back. A bit of a shame, but it’s clear the author is much better at it.
Slow idea rhythm
If Thompson were a musician I’d be annoyed at him standing there idly jamming over some chord for measure after measure. For someone like me who likes music where the chord changes often (like most baroque music where harmonic changes come on every beat) it feels awfully static and dull. There’s no momentum, no steady harmonic movement that propels the musical narrative forwards.
Transferring the notion of harmonic rhythm from music to prose, I’ll say that Hit Makers frustrates me by its slow idea rhythm. “Idea rhythm” being the speed with which the text goes from one idea to another. Another way to phrase it is the frequency of “buts” — of subversions, course corrections and thesis/antithesis/synthesis cycles. There aren’t many buts here.
I think there might be enough similarity between a piece of music moving from chord to chord and a piece of prose moving from thought to thought that a preference in one carries over to the the other. Maybe there’s something to the Myers-Briggs distinction between “intuitive” abstract thinkers who focus on structure and “sensory” concrete thinkers who focus on concrete detail that manifests as this taste difference.
Of course the only example I have is myself, so who knows? I test off the chart as an “intuitive” over “sensory” type and I’m bored by “vibey”, static music that doesn’t skip through the harmonic landscape like a skilled parkourist. But clearly others aren’t. And equally clearly many people enjoy other features of books than twisting sequences of ideas released in well-engineered succession — features like interesting stories and neat facts.
And facts there are aplenty. Thompson makes up for lack of theoretical complexity and scientific weight with a deluge of empirics. There are lots of references to numbers, names, dates, YouTube video hit counts, event minutiae, myth and history that, like the stories, are more than what’s needed to bring the points home. I suspect their function is to supply the heft and seriousness, because I get a “school project” feel from it. It appears designed to demonstrate that the author has spared no effort in doing his research and can write really well, but not that he’s genuinely passionate about understanding his question or has thought about it in greath depth.
As a result the book lacks an intellectual vision. It’s just highly competent in a slick, oversocialized way where little true personality comes out. If the book were a person it’d be like many of the people I met at the elite business school where I briefly studied: smooth, well-oiled bundles of overly appropriate behaviors seemingly dictated by the situation rather than emerging from a strong underlying personality.
Now, I don’t dislike the book for what it is and I don’t dislike Thompson. He seems to be exactly the thoroughly nice guy the smile on the back cover suggests he is. But it’s sad that I, as the reader of his book, don’t feel I really get to hear him think. It’s all “presentation” — school project, remember? — too much like he’s speaking in front of some powerpoint slides and too little like we’re having a one-on-one conversation in the small hours after too much wine. I detect in it that quality I usually refer to as “boring competence” that seems all too common in our optimized, data-driven Netflix Era: be awesome at painting inside the lines.
Part of it is Thompson’s lack of focus on contrast, tension and contradiction (i.e. few “buts”). One such contrast with his own approach comes up when he visits mathematician Duncan Watts. Watts works on popularity and “hits” as well, and his models stress underlying structural features that lay the groundwork for popularity, and then whatever “ignites” the system is largely random. Thompson recognizes that this goes against his own approach, which assumes you can learn something by studying the success stories themselves. There’s tension here (a deep, important tension of a kind that goes back to the primordial Two Cultures) and a more thorough exploration could’ve adressed it — perhaps even used is as a foundation — rather than just mention it as a curiosity before simply moving on, undeterred
All in all, discussion of difficulty, uncertainty, and evidence pointing in different directions in a way that necessitates analysis and perhaps some speculation (it’s not a dirty word) is almost wholly missing from the book (any difficulty just leads to a kind of “it’s impossible to know” shrugging). It tends to go from facts and stories straight to conclusions, bypassing the whole layer that turns facts into conclusions — theory.
It’s no accident Thompson and Vanderbilt write like this, they’re journalists and journalism tends to select for atheoretical, easily digestable writing in a neutral, uniform style. It’s a pity and I wish we had more popular “auteur type” nonfiction writers who put a lot more personality, speculation and unorthodox form into their writings. I think part of the reason we don’t is that pop-scienceish writers tend to come from the ranks of journalism and academia, which are both cultures that make you unlearn writing with personality.
Maybe I’m just barking up the wrong tree when I’m writing thousands of words complaining about this book not being what I hoped for. I’ve thought and read a fair amount on this topic already: a few books, even the odd scientific paper, and maybe I should just accept that books you buy at a train station are for beginners looking for a relaxing read full of facts you can whip out at cocktail parties, and what I want is much harder to find. It is a basic book. Thompson explains in the beginning that he wants to dispel the naive, “sentimental” myths about hits. Those are that 1) good stuff become hits purely on their merit, and that 2) it’s the most novel, most creative things that become hits. That’s a meaningful message, I guess, but just a little bit too much like writing a book about Santa Claus not actually being real you guys. You don’t say?
I didn’t write all this to rag on the book (ok just a a little, but I enjoyed it for what it was) but because I’m fascinated by how different people can be and exactly what those differences look like. Hit Makers (as well as You May Also Like) is written by someone who doesn’t think like me, and there’s something interesting about that. And in this case the topic is trivial enough to not provoke annoyance and outrage in me, which is an additional plus.
Thompson is nothing if not consistent. Even in the very last chapter, when it seems like he’s going to tie it all together by repeating and slightly developing the main takeaways, he can’t resist going on yet another story-bender and spend several pages on a detail-packed history of the Disney corporation and yet another mini-biography, this time of a rapper. He does hint a little bit at theory and says that these two stories point to the future of hits: massive conglomerates and independent creators. But like before, I get the feeling the stories themselves lie closer to his heart than the model they’re meant to illustrate.
It’s entirely possible a sequel could be written that pulls all this together into a truly interesting whole, but as it stands it does little more than vaguely gesture in the direction of The Truth About Hits. Weirdly it makes me think of Infinite Jest, which I read about two years ago and also wrote a half-critical review of. The similarities might be smaller than the differences — Infinite Jest is as ambitious, personality-laden, unorthodox in style and far from hyperpalatable as a book can be — but they do have in common that a meaningful conclusion might exist “beyond the edge of the right frame” as David Foster Wallace said about his own book.
• • •
The contrast is especially clear to me because I also recently read the first volume of Christopher Alexander‘s The Nature of Order. Now that book has a purpose and a strong thesis that Alexander works hard to drive home. It’s a complex, abstract point about Structure that takes a lot of explication and convincing to make, so of course there are a lot of examples. Pages and pages of them. But I never get the sense that they’re there for anything other than to illustrate the ideas.
It comes as no surprise when I notice on the first page, when revisiting the book just before finishing this post, the information that several parts of it has been previously published in the Atlantic. That makes sense. It certainly has the feel of one of those books that’s been adapted from a series of loosely associated articles.
He does say this:
I’ve tried to be careful in my drafting to select stories that illustrate the most universally established and rested principles—like exposure effect—rather than go about things in the opposite way, by finding stories that are interesting and then hunting for theories to shoehorn inside them.
The story of the self-publishing rapper ends weirdly with a sharp turn (a But!) at the last page where it stresses the importance of knowing the right people — a bit of a clash with the “Empire and City State” motif that also could have formed the basis of a more but-driven book. In the end he comes to the conclusion that success is serendipity and depends at least as much, if not more, on meeting the right powerful people at the right time as it does on talent and hard work. It sounds like he intends the ending to have an inspiring quality, but it certainly doesn’t. It’s depressing in its demonstration of how weakly success and merit is linked, and connections seems like a less “fair” key to success than even random chance.