Why IIII Instead of IV on Clocks?

Timing is one of the key parts of digital system architecture and design. Humans have been putting significant effort in to the design of methods and devices to keep track of time from our beginning. Awareness of and ability to measure time is one of the key differentiators of intelligence that separates humans from animals. I’m always amazed as a computer engineer that I talk as if 100 ps is an eternity, and yet my human brain has no concept of this duration of time. When I first started working as a digital engineer 23 years ago, I would have said 10 ns instead of 100 ps in that previous sentence. I suppose that an underlying fascination with time measurement is one of the things that lead me to an interest in computers, then to Computer Engineering, High Speed Digital Design, and now Systems Engineering.

About 23 years ago, as I started my career, I became fixated on the question of the title of this article. It seems in recent years it is much more common to find Roman numeral clocks that have IV on them than it used to be. At any rate, while the internet and Google existed at the time, the answer to this question was not to be found with online searching. So I spent some time in the Appleton Public Library looking for the answer. Recently, I was going through a box of my old stuff and came across some photocopies of some references that I found in my research.

In [1] on page 69 there is a description of a specific type of watch. There is a comment about the shape and size of the four (IIII). There is an offhanded follow up comment that from the very earliest times, IIII was used instead of IV because the IIII is heavier in appearance and better balances the VIII across from it on the dial. It was an aesthetic reason.

In [2] on page 105 there is a description of the dials of lantern clocks. The fact that Roman numerals were almost always used and that the more ancient version of the four was used (IIII) even though the IV was the accepted Roman numeral representation of four at the time. It further notes that this usage persisted almost exclusively until the end of the 18th century. I remember reading that the Roman numeral system originally used the four marks for the fours and nines instead of the single pre-mark to denote one less and that it was changed to be more efficient at some point, though I don’t have a reference to that offhand. The fact that it was at some point the normal way to represent a four, lends credibility to the idea that clock makers didn’t just make up IIII for aesthetic reasons, but that they chose it from available options for four because it looked better. The fact that they chose to represent the four with four marks notation and the nine with 1 mark notation clearly denotes that it was not engineer or mathematician who made the decision but an artist!

In [3] on page 61 there is further confirmation of the aesthetic virtue of IIII over IV in a description of the long case clock. There is however an interesting anecdote mentioned that gives another possible origin of the IIII instead of the IV. Henri de Vick, an early French clockmaker, created a clock for his king that had IV for four. The king declared that the IV was wrong and should be IIII instead. De Vick insisted that he was correct, but was promptly reminded that the king is never wrong, and he thus changed it to IIII.

In [4] on page 96 there is description of a clock design from the late 17th century that used the IV instead of the more customary IIII. It then gives the reason for IIII in common use as the aesthetic one as mentioned in the other references. One interesting thing to note that the reference points out, is that if you ask most people to write the digits of a Roman numeral clock from memory without looking at one, almost everyone will write the four as IV, though clearly many of such clocks that they have seen in their life will have had IIII. Perhaps it’s true that not many would pick up on this subtle difference in clocks, fewer would care, and that only an engineer would spend hours researching the reason for it at a library, and write a blog article about that research 20+ years after doing it. Mostly likely only a computer engineer!

[1] Clutton, Cecil and George Daniels. 1965. Watches. Viking Press, Inc.: New York.

[2] Wenham, Edward. 1964. Old Clocks. Spring Books: London.

[3] Bruton, Eric. 1968. The Longcase Clock. Frederick A. Praeger, Inc. Publishers: New York.

[4] Bird, Anthony. 1973. Illustrated Guide to House Clocks. Arco Publishing Company, Inc.: New York.