Why Clocks Use IIII Instead of IV

Museum catalogues and official sources show both forms on real clocks.

• IIII on a sixteenth-century clock. The British Museum's astrolabic table clock attributed to Jeremias Metzger, from around 1580–1600, has a hole "between IIII and V of the main chapterring" for releasing the hour-strike, according to its catalogue entry.
• IIII on quarter dials. A chamber clock made by Nicholas Vallin in London in 1598 is marked "I - IIII for the quarters" (British Museum), and an astronomical clock from southern Germany, about 1620, has quarters marked "I, II, III, IIII" (V&A).
• IV on the most famous clock in London. The UK Parliament notes that on the Great Clock of Westminster, Big Ben's clock, "the hour figure of 4 o'clock is shown by the Roman numeral IV, rather than the usual IIII on other clocks" (UK Parliament).

So IIII is centuries old on dials, and IV appears on at least one very prominent one. Both are in use.

None of these has been settled by documents. Each is a theory, and some are stronger than others.

It was simply the older form. In Roman inscriptions the additive forms were common. Keith Houston, citing the Oxford Classical Dictionary, writes that "the additive notation was preferred for inscriptions" (Shady Characters). On this view, clockmakers kept a form that was normal when the tradition began, while writing elsewhere moved on to IV. Mental Floss presents the same idea: clockmakers "hung on to the traditional IIII".

Visual balance with VIII. W. I. Milham, in Time & Timekeepers (1923), argued that VIII is the heaviest numeral on the dial and that a light IV opposite it would look unbalanced, while IIII matches it. His argument is quoted in the Electric Time Company's FAQ on the question, which also gives a similar view from Joerg Haus.

Casting the numerals. Ray Mialki, quoted in the same FAQ, points out that with IIII a dial needs 20 I's, 4 V's and 4 X's, numbers that suit making the numerals in sets from moulds. This explains a convenience, not a historical decision.

Jupiter. Another story says Romans avoided IV because it begins the god's name, IVPPITER. The FAQ attributes this account to correspondents who themselves ask for scholarly confirmation, and Mental Floss notes that it is undercut by the fact that IV became common long after Rome fell.

A king's preference. The tale that Louis XIV ordered IIII is told about other monarchs too. Milham himself called it probably just a story, since IIII on clocks predates the king (Electric Time FAQ).

Watchmakers give the shorter version: Citizen's support page says the choice is about tradition and the look of the dial (Citizen).

Away from dials, IIII belongs to old writing. Houston describes a Roman altar on which a cohort number appears as "coh(ortis) IIII", and the guide to the history of Roman numerals explains why subtraction was not a fixed rule in antiquity. In modern rules, IIII is flagged as invalid because a symbol can't repeat four times in a row.

• Designing a clock or watch face: either form is fine. IIII follows the long dial tradition; IV matches the modern rule.
• Anything else (a date, a chapter, a tattoo, a logo): use IV. A reader, a teacher or an online validator will treat IIII as an error outside a dial.
• Learning the numbers 1 to 20: the chart from 1 to 20 uses the standard forms only.