Wednesday, November 11, 2015

Computer Music in 1949?

Some time ago, it was revealed here on Ancient Wonderworld that CSIRAC was not the first machine to play computer music, and that actually Frances E. "Betty" Holberton programmed the UNIVAC I to perform music to the public a couple of months prior to the Sidney events in 1951.

Ever since, I have been wondering if there are perhaps even earlier examples of computer music. There are some claims that US Air Force Lt. Herb Finney wrote a music routine for the UNIVAC I serial #2 before B. Holberton did hers for the #1. Unfortunately, I have not been able to find any substatial information backing up these claims. Furthermore, considering that the performance by B. Holberton might have been already in March 1951 rather than June as previously assumed, we cannot be sure who was first based on the currently known facts.

Then some time ago I stumbled across this odd CYHIST newsgroup message from 1997, wherein a certain Kathryn Kleiman claims that Betty Holberton told her that

"[...] in the spring of 1948, on the eve of the introduction of the BINAC, a party for all the employees of Eckert-Mauchly Computer Corporation was thrown. As a highlight of this party, she programmed BINAC to play "For He's A Jolly Good Fellow" for all the employees who had worked so hard to get the machine out the door. The performance was quite a hit!"

Now that would be quite amazing. Especially considering that the BINAC only ran it's first test program on February 7th, 1949, and certainly wasn't operational in 1948. So until now I had good reason to believe that K. Kleiman's claims are incorrect, and that she got confused with the events surrounding the UNIVAC I. There is evidence that Louis D. Wilson installed a diagnostic speaker on the BINAC; but that's about it.


Or so I though. Until I came across this fascinating oral history interview with Jean J. Bartik and Betty Holberton, conducted by Henry Tropp in April 1973. On to page 169:

BARTIK: Well, we had it singing songs and stuff and playing music.

HOLBERTON: Well, that was done at the demonstration because that I did the night before the demonstration, and I don't remember when —

BARTIK: And when was the demonstration?

HOLBERTON: When was the demonstration in relation to shipment?

TROPP: That would have been fairly late

HOLBERTON: Fairly late.

TROPP: in '49.

HOLBERTON: It would have been in the spring. Yes.

Based on this, it is safe to say the above statement by K. Kleiman is indeed true, save for the year. I have not been able to pinpoint the date with certainty, most likely they are talking about the official verification test, which took place on April 7th, 1949.

Betty Holberton at the ENIAC

In any case, there you have it: The world's first computer music was made as early as 1949. So next time you write something about Betty Holberton (then still known as Betty Snyder), don't just mention her pioneering programming work on the ENIAC, the writing of the first computer manual, the development of the C-10 and COBOL programming languages, or the design of the UNIVAC console. Also mention that she invented computer music.

Sadly, no recordings were made of that historic achievement. Perhaps the papers by Betty Holberton contain some information about the inner workings of the music routine at least, so if anybody happens to stop by the Charles Babbage Institute in Minneapolis, I urge you to take a look at them and let me know about any possible findings.

Also, if you do have any other stories about early computer music, please do share them in the comments below.


  1. Earlier than Peter Zinovieff's experiments! I wonder if he knew about this when he started

    1. Good question! I assume he knew that this kind of experiments had been done (as it was actually a pretty common thing to do in the mainframe era), though he might not have been aware of these particular events.