All Hail The Mighty Octothorpe


You used an octothorpe last week. You’ll use one again this week. Maybe you’ll see a tweet with an octothorpe inviting you to a webinar, then when you log in, you’ll dial a phone number and enter your entrance code using an octothorpe.

Figured it out? It’s the name that Bell Laboratories engineers applied to the tic-tac-toe crosshatch symbol: #

The derivation and use of this humble little symbol is fascinating. Every few years, someone happens across the trivia associated with it and breathlessly writes up an article about it. This week it’s my turn. Of course, as with all other human knowledge (both factual and apocryphal), it’s now available on Wikipedia with footnotes. So go look it up if you want more details. You can have a lot of fun following connected facts that will lead you from Isaac Newton to early telephone systems to ASCII codes to typewriters, bookkeeping, and accounting.

It all starts with people writing about weights in scientific texts. The preferred terminology was the old Latin phrase “libra pondo” for “pound weight.” That’s a lot to write out every time you want to indicate a weight in an experiment, so they created an abbreviation. Instead of the perfectly obvious “lp,” some genius decided on “lb” instead. Don’t ask me why.

There was a convention of adding a horizontal crossbar on the abbreviation. I have read one account saying this was a common way of connecting two letters to show that they were an abbreviated form of something longer. Another account says it was merely a way to clarify that the lowercase L was not a number 1 or an uppercase letter I. Our writing system has a lot to answer for in terms of disambiguation.

Anywho, when you are writing that symbol over and over again in your logbooks and journals, you can get a little lazy and not bother lifting the pen off the paper to make everything perfect. Here’s how Isaac Newton wrote the libra pondo abbreviation somewhere around the late 1600s or early 1700s:

Isaac Newton's abbreviation for libra pondo

Aha! Tighten up those vertical loops, flatten out that rising line that goes up to the crossbar, and squint. We’re approaching the general shape of a crosshatch. With more time and laziness applied to the evolutionary process, people eventually started just writing the two vertical and two horizontal lines as the abbreviation, often forgetting where it came from in the first place.

This is why it’s funny when people who grew up with a UK-based English language education get so hot and bothered by Americans calling it a pound symbol. It literally is the symbol for pounds as a weight measure. Of course, the English language being such a mess, the Brits had to go and use the same word for their currency, and then the ₤ symbol took over primary duties for indicating pounds sterling. So they had two different symbols competing for use of the same terminology. Money always wins in those situations. Americans avoided that little conflict by inventing the dollar with its own currency symbol, so we were free to continue using the original “pound symbol” designation for the crosshatch abbreviation.

Now we come to the strange question of how the symbol got conscripted into use for designating numbers. I have not found an authoritative answer. But I can make an educated guess. People were doing bookkeeping for many years with the # symbol indicating how many pounds of an item were bought or sold or shipped or stored. If you start keeping track of some goods in one-pound units, the number of pounds equals the number of units. It wouldn’t be hard to start thinking of the # symbol as a way to indicate how many of something were being recorded. And voila! Two competing use cases.

By 1917, a secondary school manual in the US was advising students that the # symbol indicates a “number (written before a figure) and pound (written after a figure).”

Enter the telephone. I’ll assume you know that phones didn’t always work like they do today. If you are young enough to not remember a time before handheld personal multifunction devices, you might want to take a trip through the history of voice-only communication, operator-assisted switchboards, and rotary dials that emitted clicks or electrical pulses on the line. That’s not immediately relevant to our story.

What is relevant is that by the 1950s, telephone numbers were getting longer. In the US, dialing three digits for an area code, followed by a seven-digit phone number was time consuming and exhausting on a rotary phone. Engineers knew that push buttons would be much more efficient if they could figure out a way to make them practical and cost effective. While most members of the development team were working on switch technology and pulse-code modulation, a small group led by John Elias Karlin started thinking about the way that humans would interact with the buttons. They did quite a few studies (check out the arrangement alternatives they considered, taken from an excellent article by Francesco Bertelli).


The 3-3-1 format was the most practical on saving space, and that was the main reason it was selected. The advantage over the other formats was negligible (and others might have done slightly better!). The same is true for why they flipped the design from calculator mode (with 9 at the top right) to having 3 at the top right. We know it was not because they wanted to put alphabet letters on the keys… That came later. It seems that people who had NOT used calculating machines preferred seeing the 1 at the top, while accountants got confused. The non-accountants won.

Whatever the case, the engineers were now left with two blank spots on either side of the 0 down in the last row. The first push button phones had nothing in those spots. But nature and engineers abhor a vacuum, especially when it ruins a lovely symmetry and consistent design context. So they decided to add a couple of extra buttons to fill out the grid.

We’re in the period of the 1960s at this point, and computers were coming into use in large businesses (and the space program). Forward-thinking engineers concluded that people might someday use telephones to enter numbers into computers remotely. What if we gave them a couple of symbols to use? But what symbols? It was somewhat arbitrary, but polling seemed to indicate a slight preference among business users for the # symbol (which they were used to seeing in logs and reports and scientific use cases) and the asterisk (which they were used to seeing as a footnote symbol). Bingo! The modern telephone layout was born.

But wait… The Bell engineers realized they had one last problem. Terminology. The same old confusion that had existed for years was now spoiling the beauty and usability of their lovely invention. What instruction and education were they to give users of the new keypads? “Press the number button?” (But they’re ALL number buttons!) “Press the pound key?” “Pound the pound key?”

We’re never going to know the real story. There are at least three sworn and competing histories, and I have a strong suspicion that copious amounts of alcohol were involved but left out of the formal narrative. Don MacPherson says he combined “octo” for the eight lines sticking out of the crosshatch and “thorpe” because he liked the athlete Jim Thorpe. This is a little weird, as Jim Thorpe’s time had come and gone many years earlier. Howard Eby and Lauren Asplund claim that they coined the term originally as octotherp, with the trailing syllable made up as a gag to stand for the “th” sound that other languages find hard to pronounce.

Doug Kerr (who worked on the design team for the keypad) writes that his friends John Schaak and Herbert Uthlaut came up with the “octatherp” designation (note the “a” in the middle!) purely as an inside joke meant to chide the team for throwing out some good suggestions earlier. They wanted a term that was obviously terrible and impractical. Kerr went along with the gag and would include a footnote in his memos to the field when he would mention the # key (“sometimes called octatherp.”)

They were all amused when other publications started picking up the term as if it were a real word, often changing the spelling or providing their own made up derivations, such as “therp” being a corruption of the German word “dorf” for village. He says many publications changed the prefix from octa to octo just so they could claim it came from the eight prongs of the crosshatch.

Whichever account you believe, it is definitely a made up word, with no real linguistic meaning. But for some reason it has stuck around. I most often see it as written in my headline, but if you want to spell it without the trailing “e” or in its original “octa” form, I won’t argue.

So the next time you feel like getting into an international argument in some comment thread about whether the # symbol should be called a number sign, pound sign, crosshatch, hash mark, or hashtag – I suggest you buy a round of drinks for everyone involved and lift a glass in glorious unified appreciation of the octothorpe!


Online enterprenuer. Lean leadership consultant.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.