Playlisting

2017. At a workstation on the top floor of a suburban office complex, I begin assembling a playlist of songs released in 1982. I did it as a lark, mostly. Something to keep me from dipping to the parking garage on my breaks. I'm turning 35 in a couple weeks.

2012. I've spent the summer on a couch, penned in by Steel Reserve tallboys. When my friends ask, I'm still a part-time copywriter, but I'm mostly scrolling through Tumblr when not checking the couch for quarters. We're having a bunch of folks over for my 30th, and I figure it'd be neat to soundtrack the night with '82 jams: boogie and NWOBHM and rap.

I spend what feels like weeks at the 24-hour coffee shop down the road, researching candidates and tossing everything into a Spotify playlist. The party ends with me downing the remnants of several abandoned plastic cups or commandeering my own aux to play Class Actress, I forget which.

2013. I’m taking meetings with the nephew of a prominent New Age scammer, trying to land a copywriting gig. For me, this is personal progress. That same month, Michael Daddino posts a Spotify mix to the ILM forum (a place I find incomprehensible, though I’m friendly with many of its regulars). It’s an exquisite audio survey of 1950: not a best-of, not a chart countdown. And it’s fortified with non-Anglophonic cuts: cheery pop novelties, carefully inherited devotional forms, local musics strolling to the dock to see what’s come in. There are orchestral and operatic excerpts; the set closes with a field recording from an unspecified place in East Africa. He had nodded to history and the construction of history, but was not beholden to it: Patti Page’s “Tennessee Waltz” is here because it’s a marvel of technology and assemblage; Pérez Prado’s constricted, now-infamous “Mambo No. 5” is not. (Instead, he pinch-hit the rollicking, relentless “Babarabatiri.”) Altogether, the 170-song project startles me: this is how you can trace musical time. This is what Spotify is good for.

I also think the playlist’s title, The Bomb at the Heart of the Century, is cool as shit.

2014. A buddy refers me for a support job at a software company. I’m doing happy hours and shit.

Towards the close of the year, Jonathan Bogart starts doling out twenty-four hour-length modules of songs from 1977, the year he was born. Each module is organized around an oblique theme (“CLOUD.” “DRIVE.” “FIGHT.”) and is accompanied by a contemporary painting. The project leans a bit heavier on popular music than 1950, but still makes room for artsier/academic offerings from the likes of Joan La Barbara and Conlon Nancarrow. And it deploys the kind of head-clearing transitions you’d find on a carefully crafted mix CD: Chicago’s valedictory sop “Baby, What a Big Surprise”—the sonic equivalent of Brian Wilson dosing George Harrison—leads into the crisply serrated power pop of Starz’s “Cherry Baby,” the kind of snap that can force you to envision the fit.

2016. I don’t know what happened this year. I think that sometime in the first half, Eric Harvey first posts When the Demon Is At Your Door, a selection of tunes from 1974 that clocks in at a mind-boggling 29 hours. “Listen in order for optimal results,” the description reads, which cracks me up. But I would, eventually, listen in order. The mix is funky and airy and laid-back; listening to it still feels like prying open a rusted Pazz & Jop ballot box. The next year, when I post the redone ‘82 mix, I nick his titling convention.

2018. I'm dashing into a pizza place, looking for a cup of water to extinguish the burning cigarette I dropped into the wooden porch out front. The cooler was empty, so I improvised and bought a pint. I have a child now.

Now two or three beers in, I regale my friend Tom with my plan to move on from single-year playlists—to this point, I have created maybe five—to one giant '60s mix. I don't have to include all the usual suspects - if I want to leave out "A Change Is Gonna Come," I can. That sentence would be my second-biggest regret of the evening.

Los Bukis, Yo Te Necesito (1982)

With today’s playlist for 1951, I have selected and sequenced nearly 30,000 songs for 22 yearly mixes. I published the 1982 mix on my birthday; it is now exactly four years later. My first efforts, I had to scratch and scrape to find 600 candidates. With each new mix, my research methods improved - as did the source databases - and now, any playlist from 1969 on routinely features 1500 tracks, minimum. My hope is to make a mix for every year since the end of World War II. I’m averaging five and a half playlists a year, but I’m not sure if the math will hold: the easier it is to find candidates, the longer it takes to piece the mix together. Also, I’m not even sure if Spotify will exist by the time I’m 50. Hopefully I will.

The one step of the process I haven’t routinized is the first: choosing a year. At first, I gravitated toward eras I wasn’t familiar with (1989, 1955). Sometimes, I’d pick something that had a reputation as a “bad year for music” (1974, 2000). Perpetually, I’d take requests. 1966 was for a member of Weird Celtics Twitter; 1967 was for a comedian/pure soul/one-man Hollywood story factory who usually gets my updates before anyone else; 1984 was for the indefatigable Ryo Miyauchi (who is basically, heroically, playlisting the world in real time these days).

These days, my guiding principles are these: fill in the map evenly (keep the gaps down). And try to work in anniversary years when possible. In 2020, for instance, I assembled mixes for 2000 and 1960; this year I’ve wrapped 1981 and 1951, with hope to finish 1971 before Christmas. It’s not like the anniversary mixes garner more interest—each mix gets about 30-50 likes unless, like, Kurt Loder quote-tweets me—I just think of them as correctives to a culture industry that is starving for nostalgia but keeps reaching for the same bones. (To be clear, though: anyone who’s ever gotten hype for one of these mixes is a real one, and I think about them every time I have something new to share.)

Middle of the Road, “The Talk of All the U.S.A./Samson & Delilah” (1971)

Once I settle on a year, it’s time to assemble the longlist. I use four primary sources:

AcclaimedMusic, to see what critics liked. I don’t have great recall for dates, so I like to start here for the heavy hitters. The site was founded by Henrik Franzon, a Swedish statistician. He and his forum of volunteers collect critics’ lists from around the world, and he plugs them into a proprietary methodology to sort out the best-reviewed albums (since 1950) and singles (since 1920). European publications are well represented in its critical dataset, which in turn flags chansons and continental hits for me, as well as Francophone artists from Africa. Franzon maintains a large number of Spotify playlists as well.

Wikipedia, to see what the listening and/or record-buying public liked. They offer a page for the year in music for each year since 1550 (!); most 20th-century pages have a section titled “Top hits on record” or something similar, which just lists a bunch of tunes in alphabetical order. Near as I can tell, these sections tend to be fairly subjective collections, in the best sense. They’re a decent proxy for how a year is remembered, particularly for English-speaking users. Additionally, there are dozens of category pages that collect chart hits: top 20 US country hits, adult contemporary top 10s, #1 from Korea, Argentina, Mexico… I go through every page. Unless the song is a complete non-entity (or I already have something from the lead artist), it’s added to the pile.

Discogs, to see what collectors and sellers are into. Way back in 2012, the site was still focused on dance music; I’d navigate to some Mexican pop band’s page and see a slew of missing release years, or a gap where releases should have been in the first place. Since then, though, it’s become a much better database for bygone pop music from around the world. I use Discogs as the tiebreaker when the release year differs between sites. As you’d expect, its users are more interested in cataloging than classification, but if you take a broad style (like “Folk, World, & Country Music”) and sort the results by Most Wanted, you get a nice mix of the broadly renowned and the hungrily coveted.

Or I’ll run a Google site search for terms like “recorded” or “originally released,” plus the year. It’ll turn up various-artists compilations or archival entries that have notations for the individual tracks. Lately, I’ll also browse the list of countries/regions of release, starting with the smallest number of recordings. Sometimes you’ll get a Turkish pressing of Boston or the Mauritian issue of Like a Virgin; often, you’ll find Ottoman classical music or sega.

The region-of-release modal. Gulf Cooperation Council!

Finally, RateYourMusic, to see what people who are Too Online like. It’s a free tool for cataloging collections, building wishlists, posting reviews, and overrating boom-bap albums. If you’ve ever seen someone refer to Sweet Trip or 100 gecs or (if you’re my age) Maudlin of the Well as rymcore, that’s the source. Seattleite Hossein Sharifi created the site in 2000; I’ve been using it since 2005. In the last few years, Sharifi has built a robust query functionality that allows users to create charts based on user ratings. Best power pop EPs of the ‘80s? Top singles released by artists hailing from Kinshasa? The highest-rated esoteric EBM albums from 1990? I think I’ve run all of those queries at some point. Even sorting by the worst-rated releases is worthwhile: a negative approach way to discern popularity (or notoriety). This is the bulk of my work: building charts, playing candidates in Spotify, and evaluating songs. Usually, I’ll audition the tracks with the highest average rating—I pay a subscription fee to see this info—or I’ll just go straight for the most interesting title. I don’t dither. If it lands any kind of punch, it’s in. And always, always, I’m giving the benefit of the doubt.

Claudette et Ti Pierre, Tour 79 (1979)

July 2020. Everyone’s asleep; screen glow floods my side of the bed. I’ve got Twitter in one tab, the Washington Post in the other: video footage of a white supremacist police state in panic alternates with strangers’ GoFundMe links and bail info. Charts and databases in the other tabs. I jump between all of them: feeling small and still and halfway through my lifespan. I’ve dropped every ‘81 rap release I can find into the longlist. I set the laptop on the floor and stagger to the fridge for another Double Digit. Moving on to heavy metal.

March 2020. Time was, if the rest of the family was elsewhere, I could post up at a local bar, thumb through Safari, and add tunes. It wasn’t terribly efficient, but it was a nice recharge for an hour or so. I used to build playlists like this: holding a phone to my ear, sitting on a bench or a stool, tearing through recorded music history. When I go home, it’s time to cook dinner, build magnet tile houses, draw baths. I can pick it up after bedtime.

On the 6th, I take the afternoon off (mental health day) and head to Indian Roller. The bartender is chatting with a couple customers about the news: Austin, out of what is characterized as an abundance of caution but a paucity of knowledge, is canceling SXSW. I ask the bartender if she’s concerned. She isn’t.

David Guetta & Chris Willis ft. Fergie and LMFAO, “Gettin’ Over You” (2010)

Beyond the aforementioned quartet of sites, I add whatever I bumble across: YouTube links from my Twitter timeline, the Vintage Obscura subreddit, Pazz & Jop results (including everything that received a single vote in 2010), NPR’s Turning the Tables series, this private-press list. I’ve auditioned the entire Numero Group catalog at least four times. (If a song’s not available on US Spotify, I’ll search for it on Spotify’s website and add it that way, if it turns up.)

Once I’ve banked a large enough number of tunes I’ll see what other playlist makers did for the year I’m researching. I draw massive inspiration from the painstakingly assembled Musicophilia box sets and the transportive, multimedia assemblages from James Errington. A couple years back, I happened on Ryan Dee’s 1979 playlist. Ryan’s single-year playlists—starting with 1977 and steadily advancing—cohere like mine never could: jazz-funk and post-punk and dub and downtown skronk speaking their peace. Investigating Ryan’s work spurred me to make better use of archival sources; his Spotify profile is dotted with playlists devoted to incredible archival and reissue finds, grouped by year.

A note, then, about eligibility. A song is eligible for a given year’s playlist if:

1. It was first made commercially available in that year.

2. It was recorded that year, but was not made commercially available until three or more years afterward.

I have no interest in researching when a given single was first marketed to pop radio, or peaked on the Hot 100. People make Billboard-specific playlists for that. To me, Sheriff’s "When I'm with You" is a 1983 song: I’m cataloging the work of a given year more than its successes. Also, as I dip toes into the post-p2p years, I’m slowly realizing that white-label issues and online premieres are just as valid as commercial release…

Each lead configuration can only appear once. So Frank Sinatra appears on 1951 twice: once with the existentially pining “I’m a Fool to Want You,” and once with the brutal novelty duet “Mama Will Bark,” with Dagmar. (Incredibly, both songs appeared on the same single. “Mama Will Bark” was the A-side.) There is no limit to features (key when adding filmi cuts, or documenting the rise and fall of hot rappers across the years), and I’m pretty tolerant of aliases as well: I think Joseph Longo appears on the 1990 mix three times.

There are also bad people here: bigots and predators and reactionaries and abusers. You don’t need examples: we know who many of them are, and will continue to find them out. But they are still part of a given year’s story, and so I usually include them in the mixes. (For now.) Still, I hope Spotify’s block-artist feature is providing use to some listeners.

It takes me about two or three months to gather the raw material. Eventually, I know when I’m done: the ratio of new finds to the already-evaluated tips too far. Then it’s on to the mix. I use a tool called Exportify to produce a CSV of track data. I then paste the Artist and Song columns into a Google spreadsheet, remove duplicates, then add three more columns:

Whenever I drop a song into the mix, I add a character to the Added? column. That way, I can filter out cells with values. I used to use an x, now it’s a 1. For 2010—after Spotify’s latest garbage UI update, I thought I was condemned to interminable scrolling in order to locate songs within a mix—I created mini-playlists, capped at 140 tracks, each named for a different letter which I would put in the Added? column. I eventually learned that to jump to the current song’s place in a playlist, you just click on the cover art (as long as the playlist isn’t currently filtered). I stuck with the mini-playlists, but I dunno if I’d do that again.

The Descriptors column is where I list stylistic and emotional descriptors for each track. Filling out this column can take as long as the mixing, but I like what it does for the transitions. There’s a sweet spot between terse and flowery: the goal is to look at a given cell (rap crew handoff eminem upstroke, downtempo new age wistful, r&b neo soul skittering percussion) and get a general sense of what the track sounds like. For 1951, once i placed Webb Pierce’s “Wondering,” I could filter for honky tonk and search for every resulting cell that contains “ballad”.

Notes is where I add streaming links for songs not playable on US Spotify, or suggestions to myself for placing a song towards the beginning or end, or maybe for pairing it with some other song.

(If you have any suggestions for improving this process, I am always interested.)

The spreadsheet for the 1981 playlist, filtered for “ska” descriptors.

Like a true psychopath, I manually sequence every song in every playlist. If I were making a mixtape for a friend, I’d want to keep switching the mood up. Here, I’m going for pure flow: as much as possible, I want legible transitions, to let you drift through a year. My hope is that if someone particularly likes a song, they’ll also dig the songs before and after it, and maybe the songs before and after those. This is where the Genre column comes in handy. I love the songs that straddle styles, the experiments and hybrids. Songs like Floyd Dixon’s “Do I Love You,” which bares its R&B heart across a European cafe table. Or Underoath’s “Paper Lung,” which gives you Deftones-style modern rock angst before forcing the metalcore medicine. People hate reggae cross-pollination, but for playlisting purposes, it’s nearly always a tropical delight.

When mixing, I have a few guidelines:

• Consecutive tracks should not feature the same artist. Producers are fine.

• Try not to sequence too many big names or chart hits in a row.

• Try not to have interminable stretches of one genre. I break this rule a lot. I routinely offer up 30-track stretches of death metal or Southern soul: good news if you like (or wish to explore) either genres. Grim if you’re into other things. Keep going.

In this way, I kind of through-compose the playlist: once I dispense about two-thirds of the tracks, I start figuring out how it’s going to end. Then I start sprinkling the last third into the mix. Then I title the mix after a lyrical fragment—again, a concept I shamelessly ripped off of Eric—and give it a photo of a musician. (The photo has to be taken that year.) Then I post it for the 10 or so diehards.

Raffaella Carrà, “Rumore” (1974)

With each mix, I have two primary goals. The first is to find as many songs as I possibly can, from as many musicians and styles and countries as I can encounter. The second is to stitch these finds together in a way that’s a lot of fun—legible, hopefully, but mostly fun. I know of a couple people who listen to these straight through–I mean, I know they’re skipping some tracks and bailing on others, I’m not kidding myself, sometimes I’ve dropped 60-minute noise tracks like landmines—and a couple more who have used a mix as temporal background noise for historical research. I know of many more who haven’t gone near these things (or don’t fuck with streaming), but are nonetheless happy they exist. That’s cool.

In a just world, I wouldn’t be using a streaming service at all, unless it—and I—were paying our fair share to those who labor fuels it. I agree with most of the other criticisms of streaming as well—you don’t own anything; the databases are littered with incomplete, misspelled or incorrect entries; you can’t read liner notes or properly view cover art; the UI forces each recording into the same bland, minimal presentation; there are hundreds of albums on these services that were recorded by a child holding a tape recorder to a thrift-store LP.

When it comes to their in-house playlists, though, I’m agnostic. I’m not going to wring my hands about someone’s streaming experience being mediated by a corporation; if Apple Music or Spotify ceased to exist tomorrow, that person would find a new mediator. And I would probably switch to making box sets.

In the meantime, I keep at it. Creating these mixes has made me a better listener: I truly believe that. There are those who recoil at the idea of new age, or atmospheric black metal, or adult alternative, so when they engage with it, there’s no illumination, only the sound of a shorting fuse. I don’t love every one of these 30,000 songs. But they were made by people: schemers, tinkerers, idealists, cynics, sellouts. They were made by us. I’d like to remember them until I run out of years.

Cloud One, Atmosphere Strut (1976)