A few weeks ago, I gave up on Serial, the This American Life spin-off podcast. I realize that my not listening to something is hardly news. But three times in the past couple weeks I’ve had to explain to friends (from three different periods of my life, no less) why, despite being a TAL fanatic, I’m no longer interested in Sarah Koenig’s weekly musings on whether Adnan Syed did or did not kill his ex-girlfriend, Hae-Min Lee, in 1999.
I love TAL. I’ve listened to a large majority of the show’s episodes, and I’m currently going through them from the beginning and listening to every one I haven’t listened to before. (A recent instant favorite: “Throwing the First Punch,” with three equally fantastic acts.) One of my proudest moments was getting a “special shout out” at the end of a TAL episode back in 2009, when I was at my blogging peak.
While TAL has a prismatic, innocent-eye, barely-structured-around-a-common-theme format, Serial is almost exactly the opposite — a whodunit that is relentlessly focused on one question: Did he do it, or not? This means that, for the show to sustain your interest, you have to believe that Sarah Koenig, the show’s host, and the not-inconsequential resources of the TAL team can actually help answer that question.
I don’t deny that there is some value to the investigative work that Koenig is doing, particularly when it comes to looking through the detectives’ file to find evidence that they may have overlooked, consciously or not. But too much airtime — particularly airtime that is supposed to have significance — is dedicated to fifteen-year-old memories that are, frankly, worthless.
At the beginning of episode 9, Koenig focuses on two pieces of “evidence” that were called in recently. In the first, Laura says that, in 1999, there were no pay phones at the Best Buy; this matters because Jay, the state’s star witness, said that Adnan called him from a Best Buy pay phone and that Adnan was standing next to the phone booth when Jay picked him up. “You’re sure,” Koenig asks. “I’m positive,” Laura responds.
In the second, Summer says that the state’s theory — that Hae was dead by 2:36 PM — is impossible. Summer and Hae were co-managers of the high school wrestling team, which had an away match that day. Summer says that, after school, Hae told her that she wasn’t going to the match on the bus. They had a conversation that began after the regular buses had cleared the school lot and lasted at least ten minutes, which makes it virtually impossible that Hae could be at the Best Buy and dead by 2:36. “You’re sure that this is the day because it’s the day she didn’t show up?” Koenig asks. “I am positive. I’m very positive,” Summer says. Compared to another witness who said contradictory things about whether Hae was planning to take the bus, Koenig concludes, “I trust Summer’s memory more, and Summer is clear: Hae told her she was going to drive herself there.”
The problem is that fifteen-year-old recollections like this are worse than worthless: they’re pernicious, because they present themselves as truth.
Human beings just don’t have the ability to unerringly recall mundane events from the distant past. That’s not the way memory works. The brain does not have one set of neurons that store information and another set of neurons that recall that information the way a computer can copy data from hard drive to RAM. Instead, it is impossible to recall a memory — to think about it — without modifying it. The implications of this neurological fact are clear: the memories that you think about the most, and that you are therefore most confident about, are the ones that are most likely to be inaccurate. They are “stronger” in the sense that, because you think about them frequently, you are unlikely to completely forget them. (Check out the last act of this TAL episode for a poetic example.) But they are also highly untrustworthy. (Conversely, the most accurate memories you will ever have are when you suddenly remember something you haven’t thought about since the day it happened—as with Proust’s madeleine.)
We also know that memories are extremely unstable. Scientists can implant false memories in people remarkably easily; those people typically take a suggested false memory and embroider it with real facts, creating a believable whole. This is part of a class of common memory errors. The underlying issue is that the human brain is very good at creating associations between different things, but doesn’t have a way to turn this power off when it wants to preserve some information in an undistorted form. I’m not saying that memories are all wrong, or that they’re worthless. I have plenty of memories from fifteen years ago, like eating pizza at the Cheese Board or harvesting lemons from the tree in our yard in the Oakland hills or ordering lunch for the McKinsey team at our client’s offices in Mountain View. But I couldn’t tie any of those memories to a particular day without some kind of reliable written record.
So I have no doubt that there was a day when Hae told Summer she wasn’t going to take the bus to an away match, and Summer was mad at Hae because Hae was more experienced and Summer needed Hae’s help. But was it the day that Hae disappeared? Maybe. But maybe Summer is now, fifteen years later, conflating two days in her head. The fact that she had been listening to Serial is not irrelevant: our memories suffer from suggestibility, and Koenig’s earlier speculation about where Hae was immediately after school could easily have slipped into Summer’s brain and fused with her memory of being ditched at a different match. We know for a fact that Hae did ditch Summer that fateful day—she disappeared, after all — but maybe Hae told Summer during school, not after school, or maybe she didn’t tell her at all.
What about the fact that Summer — and Laura, for that matter, to whom all these observations apply equally well — is completely confident about her memories? As Elizabeth Loftus, one of the world’s leading experts on false and distorted memories, says, “just because someone says something confidently doesn’t mean it’s true.” That was in the context of eyewitness testimony, which is usually taken relatively soon after an incident. The scientific evidence about the unreliability of eyewitnesses accounts — and the ease with which they are influenced, intentionally or unintentionally, by law enforcement personnel — is massive. Then add fifteen years.
At the end of the day, these are questions about probabilities. Given what we already know about Hae’s whereabouts, and given the risk of inaccuracy in Summer’s memories, how much do those recollections increase the probability that Hae was actually at school when the police say she was being murdered by Adnan? A tiny bit, maybe. But certainly not enough to warrant the long pause followed by the ominous music, heavy with implication, that Serial drops on us for dramatic effect. (Start listening at around 8:55 to get a sense for the scoring.)
Ultimately, that’s what doesn’t work for me: the attempt to manufacture drama.
As I said above, there is some evidentiary value to the work that Koenig does digging through old case files and finding a potential alibi witness or cell phone records that don’t match the detectives’ timeline. There could be some legal value, too. As a law professor (and someone who is generally sympathetic to defendants), I found myself listening for two things: ineffective assistance of counsel and failure to turn over exculpatory evidence. Those are Adnan’s best chance at overturning a conviction that happened fourteen years ago. Raising doubt about his actual guilt or innocence is pretty much irrelevant at this stage. Serial touches on the ineffective assistance issue, but only very lightly. That I totally understand. Most criminal appeals actually make pretty boring stories: you focus on a small number of facts (e.g., the failure to talk to Asia, Adnan’s potential alibi witness) and wrap them in the strongest legal argument you can make. That wouldn’t make good radio.
But in pursuit of good radio, Serial goes on a quest for the Truth using the storytelling method, in which you go and talk to lots of people and build a narrative out of what they have to say.
Only, in this case, most of what they have to say isn’t worth very much — and you can only make a good story if you are willing to ignore that fact.
I’ve thought a probably excessive amount about why I love TAL but I’m tired of Serial and I ultimately grew disenchanted with Planet Money, another TAL spinoff that had some wonderful shows during the financial crisis. I think it’s that trying to answer difficult questions carefully does not actually make for good radio. It isn’t particularly suspenseful, it doesn’t involve dramatic interviews with long-lost witnesses, and it can’t be done by finding illustrative slices of life from the American heartland. Some of the early TAL/Planet Money episodes during the financial crisis were great because they used relatively ordinary people to explain things that were complex but not at all controversial: what a mortgage-backed security is, for example, or how mortgage brokers made money. But when Planet Money got into trying to explain what’s wrong with the American health care system or Social Security Disability Insurance — well, those are legitimately difficult issues, and there’s a reason why some very smart people dedicate their entire lives to those topics. And the tools those people use aren’t man-on-the-street interviews — they are massive datasets and subtle statistical techniques that go far beyond cracking open a copy of Stata and running an OLS regression.
I feel roughly the same way about Serial. If you want to get Adnan out of jail, then you need to focus on the legal arguments that could actually work on post-conviction appeal. If you want to find out who killed Hae — I don’t know how you would do it at this remove, and
talking with people fifteen years later may be the best you can do, but you shouldn’t use the awesome narrative talents of America’s best radio team to invest their untrustworthy memories with unwarranted significance.
Today I was listening to “Niagara,” a TAL episode from 1998. There’s a passage, around 17 minutes in, in which a man talks about how his brother went over the falls in a barrel. The narrator says that he told his brother that his plan (involving inner tubes wrapped around the barrel) wouldn’t work. Then the he describes going with his brothers and his mother to wait at the bottom of the falls.
“I was standing on the dock at the Maid of the Mist when I seen this jumble of tubes come out, and I turned to my mother and said, ‘Bill’s dead.’”
It’s raw, brutal, and unnerving. It takes you out of wherever you are and puts you someplace else you’ve never been before. It reminded me of the most moving piece of radio I ever heard — also on TAL, of course — in which a young woman gets the results of the test that tells her whether or not she is going to get Huntington’s Disease. The runner-up is the devastating segment from “Kid Logic” in which you hear a man with a terminal brain disease losing touch with his six-year-old son.
I’ve seen Ira Glass on stage a couple of times. He likes to say that his method of telling stories is that one thing happened, and then something else happened, and then something else happened, . . . You can see that at work in all of his shows, and that’s one reason they’re hard to turn off. But at its very best, This American Life is not about narrative movement and drama. It’s about capturing and sharing slices of human experience that most people have never encountered before and will probably never encounter in their own lives.* Trying to figure out What Actually Happened to a high-school student fifteen years ago is a very different kind of endeavor — and one for which prodigious radio production skills are not particularly useful.