What is the first text considered Italian instead of Latin?

Question

What is the earliest text that is considered to be written in Italian (or a predecessor thereof), and what distinguishes it from Latin? I would like to understand the first signs of Latin evolving into Italian. (Or perhaps rather of Latin and Italian becoming separate languages?) It may be a matter of taste what is Latin and what is Italian, and therefore an answer should explain why the text should be not considered Latin anymore. A small text sample with commentary — concrete deviations from Latin — would be great.

Answer 1

According to the Handbook of Medieval Culture (Albrecht Classen, vol. 2):

The first written evidence considered to be Italian rather than Latin is known as the Placiti Cassinesi, which are four legal documents containing vernacular testimonies in an Upper Southern dialect dated to 960–963.

The text of the four placiti has been published in Storia della lingua italiana, by B. Migliorini and is available on this collection of Italian texts established by Ulrich Harsch (Bibliotheca Augustana):

I

Sao ko kelle terre, per kelle fini que ki contene, trenta anni le possette parte sancti Benedicti.
(Capua, marzo 960)

II

Sao cco kelle terre, per kelle fini que tebe monstrai, Pergoaldi foro, que ki contene, et trenta anni le possette.
(Sessa, marzo 963)

III

Kella terra, per kelle fini que bobe mostrai, sancte Marie è, et trenta anni la posset parte sancte Marie.
(Teano, luglio 963)

IV

Sao cco kelle terre, per kelle fini que tebe mostrai, trenta anni le possette parte sancte Marie.
(Teano, ottobre 963)

A philological analysis of this text is available on the Italian version of Wikipedia or in Vers les sources des langues romanes by Eugeen Roegiest.

Examples of differences with Latin Language: some forms like cco (from quod), tebe (tibi) or bobe (vobis, already betacism). However, there still are latinisms such as the use of the genitive in proper nouns or the n of monstrai.

---

Anyway, the distinction Latin-Italian was clearly established during the 10th century as a monk, Gunzo of Novare, wrote in 965:

usu nostrae vulgaris linguae, quae latinitati vicina est

An other example, according to the Manuel pratique de latin médiéval by Dag Norberg:

In 915, on the occasion of the coronation of King Berengar I, the senate paid him homage patrio ore, "in Latin," while the people honored him nativa voce, "in Italian," according to the text of a song composed a few years later.

According to Norberg, the oaths of Capoue mark the beginning of the history of Italian literature. Others claim that the oldest preserved Italian text is the Indovinello Veronese (please see Tom's answer).

Answer 2

Ostler (see my comment on Luc's answer) remarks, in relation to the appearance of a language that is recognisably Italian:

" . . . touchingly, the first surviving example of imperfect written Latin — if not yet conscious Italian — is an elegant riddle apparently used when trying out a new pen:

'se pareba boues / alba pratalia araba / & alba versorio teneba / & negro semen seminaba'

"he yoked his oxen / white fields he ploughed / & white plough he held / & black seed he sowed." The manuscript (Biblioteca Capitolare di Verona, codex lxxxix [84]) dates to the turn of the eigth-ninth centuries and is a riddle on the act of writing itself, with black ink on a white page.

Not yet Italian, of course, but perhaps of interest as approaching the date of Luc's more definite example.