How Lyrics Websites Transcribe Gibberish

YouTube
YouTube / YouTube

Beyond the ongoing misadventures of the record industry, music on the internet may be an even bigger disaster than you realized—even the simplest of indexes are in trouble. At the other end of nearly any song-related search query, scores of sites stuffed to the gills with lyrics brutally fight their way to the front of the line using the darkest tricks of search engine optimization in order to make money with a flood of cheap syndicated ads. There's little to no oversight regarding the quality and accuracy of the lyrics, since raw breadth of content tends to be more profitable in a market powered entirely by search result rankings. Design and typography are uniformly terrible. The ads are horrifically invasive. Very little of it works well on mobile devices.

Even Genius (formerly known as Rap Genius), which has won high profile supporters with a novel text annotation layer which sits on top of the lyrics—or, increasingly, other kinds of textual content—still has to resort to arcane SEO voodoo. But in many cases the source text being annotated is still the same as any of the other lesser sites, so unless you're one of the rare loons who is especially interested in the unfiltered and often inane user-submitted commentary, you're back to square one.

I. 'SCUSE ME WHILE I DISRUPT THIS GUY

Last month, Google finally started to extinguish this dumpster fire. As you've probably already noticed, there is a utility layer to Google's logic which simply tries to answer questions directly, instead of using search to route you to external sites which might provide the answer. For example, if you search for a simple math problem, Google will display the proper solution along with a handy calculator widget for further number crunching. If you add a zip code to a request for movie listings, you'll get nearby screenings, largely negating the need to check with services like Fandango.

/

In December, Google likewise began serving up the full text of lyrics to many songs, right on the search results page, before any of the links to third-party sites. Google pays the proper license fees for this privilege, which is actually a totally sensible business decision—in addition to keeping music fans addicted to their search engine, the lyrics also serve as an internal ad of sorts, pointing users toward Google's paid subscription music streaming service. In many cases, there's no longer a reason to click through to AZ Lyrics, nor even Genius.

/

It's possible that a couple of these sites may survive—Genius being the clear frontrunner simply by virtue of its tangible technological value add, and lofty ultimate goals of expanding beyond lyrics into annotations of news and legal documents—but for the most part this market is probably about to spectacularly implode. The wild west won't just be paved over, but completely erased—companies will go under, taking their servers with them, leaving no trace of all these minimally competent rival fiefdoms.

But song lyrics are important cultural tokens, and this shift would leave them represented by just one canonical reference point, which is itself served up by a giant technology company to which they are an afterthought barely worth mentioning in a board meeting. In an effort to see what else we'd lose in the process, I took a spin through all the other song lyrics sites looking for something which would be especially prone to subtle variations.

This brings us to Scatman John.

II. MEET OUR SACRIFICIAL LAMB

/

In 1994, a middle-aged mustachioed weirdo with an American Idol-worthy backstory about overcoming a socially crippling stutter scored an unlikely dance-pop one hit wonder. His self-congratulatory eponymous anthem "I'm The Scatman" layered caricatures of the scat syllables used by jazz vocalists across the cheesy Eurotrash dance tracks popular among flashpan pseudo-R&B acts of the day like La Bouche and The Real McCoy. He died just five years later, but left behind one of the weirdest relics of the '90s. "I'm The Scatman" is probably the only major pop chart hit of the past few decades to open with a completely unaccompanied scat solo.

But the Internet barely existed back in those days of Mosaic browsers and abundant AOL promo CD coasters, and bore little resemblance to its current form. Rather than instantaneous delivery from one of the most widely used products in the world, finding interpretations of Scatman John's lyrics back in the day would have likely involved sifting through an early listserv for discussing the likes of Casey Kasem, Rick Dees, and John Garabedian, or at the very least Lycosing your way to the Geocities homepage of someone who identified as a devout Scatman John fan.

But all those early technical platforms are now gone, as are Casey Kasem and his iconic radio countdown show and Scatman John himself. It isn't yet licensed by Google, so how does the modern internet treat this most bizarre of cultural artifacts when its only goal is to capitalize on it as an ad surface? On some level these snapshots seem to point toward subjective perceptions of art and the love it can inspire in us, but they are then promptly torn into nonsensical shreds by the brutal Darwinism of online marketing.

III. THE WAY WE SCABBADEEDLYBOP NOW

(Rap) Genius

Ski-bi dibby dib yo da dub dub
Yo da dub dub
Ski-bi dibby dib yo da dub dub
Yo da dub dub
(I'm the Scatman)
Ski-bi dibby dib yo da dub dub
Yo da dub dub
Ski-bi dibby dib yo da dub dub
Yo da dub dub

Ba-da-ba-da-ba-be bop bop bodda bope
Bop ba bodda bope
Be bop ba bodda bope
Bop ba bodda
Ba-da-ba-da-ba-be bop ba bodda bope
Bop ba bodda bope
Be bop ba bodda bope
Bop ba bodda bope

Ski-bi dibby dib yo da dub dub
Yo da dub dub
Ski-bi dibby dib yo da dub dub
Yo da dub dub
Ski-bi dibby dib yo da dub dub
Yo da dub dub
Ski-bi dibby dib yo da dub dub
Yo da dub dub

Genius provides an open platform for annotating text, but shockingly, none of these fascinating passages from "Scatman" have yet been annotated. Please get on this immediately, everyone; internet anthropologists of the future will thank you.

Mojim.net

B-B-B-Be Bop a Bodda Bop
Bop a Bodda Bop
Be Bop a Bodda Bop
Bop a Bodda Boop
B-B-B-Be Bop a Bodda Bop
Bop a Bodda Boop
Be Bop a Bodda Bop
Bop a Bodda Bop

Ski-Bi dibby dib yo da dub dub
Yo dab dub dub
Ski-Bi dibby dib yo da dub dub
Yo dab dub dub
Ski-Bi dibby dib yo da dub dub
Yo dab dub dub
Ski-Bi dibby dib yo da dub dub
Yo dab dub dub

The second passage is mostly the same, but we see slight changes to the first, in the form of a new initial consonant sound, and in an unclear distinction here between "bop" and "boop" upon which entire undergraduate linguistics theses could be based.

Music World

Ski-Bi dibby dib yo da dub dub
Yo dab dub dub
Ski-Bi dibby dib yo da dub dub
Yo dab dub dub

I'm the Scatman

Ski-Bi dibby dib yo da dub dub
Yo dab dub dub
Ski-Bi dibby dib yo da dub dub
Yo dab dub dub

B-B-B-Be Bop a Bodda Bop
Bop a Bodda Bop
Be Bop a Bodda Bop
Bop a Bodda Boop
B-B-B-Be Bop a Bodda Bop
Bop a Bodda Boop
Be Bop a Bodda Bop
Bop a Bodda Bop

(Scatting by Scatman John)

Having established in our last few examples the entirely sensible possibility of simply not transcribing the scat syllables at all, we can now move on to the weirdest interpretation of the bunch. Why distinguish between primary scat singing which ought to be transcribed, and the secondary phrases which don't deserve it?

Similarly, why repeat the name of the artist here? It's not as though there are multiple scat singers featured on a single song. This unusual annotation can probably be traced back to the search engine optimization industry—readers will invariably get to this page by searching for "Scatman John," not for "B-B-B-Be Bop a Bodda Bop." Reiterating the artist's name probably has little to no actual effect on the search ranking, but it's indicative of the parasitic thinking that prevails in this corner of the internet.

letras.mus.br

Ski bi di bi di do bap do bap
Do ba do bap

By now it seems that we've settled on the spelling of the first part of this phrase, but why the extra ending consonants?

AllTheLyrics.com

-skatting-
wee bababadabo bababadabo X4

Finally, a complete break with tradition! Most of our other examples thus far have seemed to be organic variants of the same source text, probably emailed out to an early Listserv back when the song was still on the charts, but this seems to be the result of hard work by a spunky newcomer looking to earn a little respect. It's a good start, but more e's seem to be in order. Further, though the scat syllables themselves are open to phonetic interpretation, the word "scat" itself is not; the unusual spelling here may be a distant result of the wild popularity of MC Skat Kat, Paula Abdul's cartoon foil from her 1989 music video for "Straight Up."

lyriczz

Ski bi di bi di do bap do
Do bam do

Bada bwi ba ba bada bo
Baba ba da bo
Bwi ba ba ba do [x2]

(More Scatting)

Here again we have that same question about primary and secondary syllables. More importantly, though, this is the only use of "bwi," which is certainly a funnier spelling than anything we've seen so far, but actually starts to seem egregiously phonetically incorrect after you've listened to the song a few hundred times, as I now have (oh god I'm really losing it here please help meeee).

LyricsFreak

(Scatting by Scatman John)

Arguably the most sensible option here is to admit that there is a logical limit to the level of nonsense that should be published on a web page. One can always hope.

LyricsMania

(Scatting di Scatman John)

Derived from the above, since SEO gaming of copyright infringement is apparently the universal language.

MetroLyrics

[omitted entirely]

This elegant solution is the preferred approach, and in fact should be used as the default treatment for not only song lyrics, but also almost everything else on the internet.

IV. SHUFFLING INTO OBLIVION

Interestingly, the parenthetical subtitle of the song is "(Ski Ba Bop Ba Dop Bop)," or in other words, there is actually a canonical partial spelling which nearly every example surveyed ignores. And there's your answer, in a way—if the internet has a constant feature which always survives across technical iterations, it is inaccuracy and ignorance. Surprise!