Northwest Semitic šîn represented a voiceless postalveolar fricative /ʃ/ (as in 'ship'). It originated most likely as a pictogram of a tooth (שנא) and represented the phoneme /ʃ/ via the acrophonic principle.
Greek did not have a /ʃ/ phoneme, so the derived Greek letter Sigma (Σ) came to represent the voiceless alveolar sibilant /s/. While the letter shape Σ continues Phoenician šîn, its name sigma is taken from the letter samekh, while the shape of samekh but name and position of šîn is continued in the xi. Within Greek, the name of sigma was influenced by its association with the Greek word σίζω (earlier *sigj-) "to hiss". The original name of the letter "sigma" may have been san, but due to the complicated early history of the Greek epichoric alphabets, "san" came to be identified as a separate letter, Ϻ. Herodotus reports that "San" was the name given by the Dorians to the same letter called "Sigma" by the Ionians.
The Western Greek alphabet used in Cumae was adopted by the Etruscans and Latins in the 7th century BC, over the following centuries developing into a range of Old Italic alphabets including the Etruscan alphabet and the early Latin alphabet. In Etruscan, the value /s/ of Greek sigma (𐌔) was maintained, while san (𐌑) represented a separate phoneme, most likely /ʃ/ (transliterated as ś). The early Latin alphabet adopted sigma, but not san, as Old Latin did not have a /ʃ/ phoneme.
The shape of Latin S arises from Greek Σ by dropping one out of the four strokes of that letter. The (angular) S-shape composed of three strokes existed as a variant of the four-stroke letter Σ already in the epigraphy in Western Greek alphabets, and the three and four strokes variants existed alongside one another in the classical Etruscan alphabet. In other Italic alphabets (Venetic, Lepontic), the letter could be represented as a zig-zagging line of any number between three and six strokes.
The Italic letter was also adopted into Elder Futhark, as Sowilō (ᛊ), and appears with four to eight strokes in the earliest runic inscriptions, but is occasionally reduced to three strokes (ᛋ) from the later 5th century, and appears regularly with three strokes in Younger Futhark.
Late medieval German script (Swabian bastarda
, dated 1496) illustrating the use of long and round s
: prieſters tochter
The minuscule form ſ, called the long s, developed in the early medieval period, within the Visigothic and Carolingian hands, with predecessors in the half-uncial and cursive scripts of Late Antiquity. It remained standard in western writing throughout the medieval period and was adopted in early printing with movable types. It existed alongside minuscule "round" or "short" s, which was at the time only used at the end of words.
In most western orthographies, the ſ gradually fell out of use during the second half of the 18th century, although it remained in occasional use into the 19th century. In Spain, the change was mainly accomplished between the years 1760 and 1766. In France, the change occurred between 1782 and 1793. Printers in the United States stopped using the long s between 1795 and 1810. In English orthography, the London printer John Bell (1745–1831) pioneered the change. His edition of Shakespeare, in 1785, was advertised with the claim that he "ventured to depart from the common mode by rejecting the long 'ſ' in favor of the round one, as being less liable to error....." The Times of London made the switch from the long to the short s with its issue of 10 September 1803. Encyclopædia Britannica's 5th edition, completed in 1817, was the last edition to use the long s.
In German orthography, long s was retained in Fraktur (Schwabacher) type as well as in standard cursive (Sütterlin) well into the 20th century, and was officially abolished in 1941. The ligature of ſs (or ſz) was retained, however, giving rise to the Eszett, ß in contemporary German orthography.