It’s unclear whether the Indus Valley symbols are a form of true writing, or an icon-type symbol set that communicates a limited range of messages. The reason some experts suspect the latter is that no inscription containing more than a couple dozen symbols has ever been found.
The Mycenaeans are known to have used at least two different writing systems (Linear B, which records Greek, and cuneiform, the diplomatic script of the ancient world, in use longer than we’ve used the alphabet), and yet very few examples have survived (and AFAIK, no literature, just records of production and religious offerings) of one, and essentially zero of the other (in any known Greek context; the archival copy of at least one letter to the Greek high king is known from Hattusas). It was easy to regard pre-Classic Greece as illiterate, which is in fact what was believed, even insisted on, until Ventris cracked Linear B in 1953.
A better example is the Etruscans — they were literate and literary, with a robust tradition; today only a handful of longer undisputed Etruscan texts survive, and at least one of those uses an adaptation of the Carthaginian alphabet, and is scratched into a metal sheet, making survival more likely. I believe the longest of the texts was preserved on a mummy wrapping, which meant it had been sold for reuse, and suggests that much of their writing material simply vanished over the centuries. The Lemnian Stele from the Aegean is apparently a dialect of Etruscan. The rest of the Etruscan inscriptions are on gravestones and are, obviously, short and largely free of any meaningful grammar. A few words and phrases survive in Latin texts from the Roman era.