From New Scientist
Perishable materials usually rot away, so the oldest string on record only dates back 30,000 years. But perforations in small stone and tooth artefacts (sic) from Neanderthal sites in France suggest the pieces were threaded on string and worn as pendants. "The wear patterns provide circumstantial evidence of early use of string, but the evidence is not definitive," says Bruce Hardy at Kenyon College in Gambier, Ohio. Similar circumstantial evidence has been found in perforated shells.

Now, Hardy and his colleagues have found slender, 0.7-millimetre-long plant fibres that are twisted together near some stone artefacts at a site in south-east France that was occupied by Neanderthals 90,000 years ago. Such fibres are not twisted together in nature, says the team, suggesting that the Neanderthals were responsible (Quaternary Science Reviews, doi.org/pzx).

From Past Horizons
As these fibres are not twisted in their natural state experiments were carried out involving incising, planing, whittling, scraping and boring. In all cases, no twisted fibres resulted.

Further experiments conducted by Bruce Hardy involved the scraping, cutting and slicing of a variety of non-woody plants (roots, tubers, reeds, etc.), and again these also produced no twisted fibres such as those observed.

While not definitive, the lack of twisted fibres in these experiments lends some credence to the hypothesis that these derive from cordage.

“If they are indeed remnants of string or cordage, then they would be the earliest direct evidence of string,” says Hardy. “Albeit very fragmentary evidence.”

"She strips the bark off the saplings she collects and peels the cambium fibre...away from the outer bark or shongol (lit. skin) which she discards. She allows the bark to dry first before stripping off its fibre, slitting the wands gathered lengthwise in two and placing them indoors over a fire, leaving them for about a week. But the fibre of ponjip figs, she may strip off green and dry separately; women often excoriate these saplings during collection, returning home with bags of unseparated bark..., and there peeling off and drying the bast. On average, 5g of dry phloem fibre will produce 10 metres of string. Sillitoe 1988 pg 245