There is a very good article in History Today by Florence Hazrat on the history (and possible future) of punctuation. She notes that:
“In classical times there were no punctuation marks or spaces between words. Since punctuation determines sense (‘Let’s eat, Grandpa’ versus ‘Let’s eat Grandpa’), scriptio continua allowed scribes to offer their masters a clean text, waiting to be interpreted by those higher up the social ladder. Writing was merely a recording of, or preparation for, speech: any punctuation that was inserted had oratorical, rather than grammatical, functions, indicating the degree of pauses upon delivery only.”
When classical texts were being rediscovered and copied in the early Middle Ages, scribes added various pauses to assist comprehension and these eventually developed into the comma, the colon, and the full stop.
“The 15th century saw a boom of inventive punctuation, including the exclamation mark, the semicolon and brackets (or parentheses). New marks arise when a lack of clarity needs to be redressed, communication controlled and sense disambiguated, an emergency perhaps stemming from greater reliance on written diplomacy as well as the newly fashionable art of letter writing.”
The semi-colon made an appearance first in 1494; while the dash and the ellipsis had to wait until the 18th century.
She concludes with a warning and a suggestion:
“When constant availability makes us minimise the effort and time we devote to messages, one may assume that punctuation is doomed. After all, December 2019 saw the demise of the Apostrophe Protection Society, because the ‘ignorance and laziness present in modern times have won’, according to its former president. Yet studies on the use of the full stop in text messaging have shown that we do care about punctuation, even in a medium that promises endless continuation. When is it time to not send another text back? A full stop, the study suggests, comes across as aggressive and cuts conversation short. Perhaps a new mark is necessary?
Jak's View of Vancouver v.3 