There are two main categories of genealogical sources, authored works and records.
1. Authored works - There are not as many authored works in Italian research then one sees in North America. An example of this might be a published family history or a town history, which may mention notable people or noble families.
2. Records - These are the battesimi, nati, morti, pubblicazioni di matrimonio, atti di matrimonio, notarial, censimento, catasto, allegati/processetti, etc. This is where the majority of evidence lies in Italian genealogical research.
Records are either original or derivative.
An original record is made at the time of the event or close to it, usually by a direct participant in the event. Therefore, a birth record reported by the midwife would definitely be an original record. Could these records still err? Of course. They were created by humans with inherent flaws and there are many reasons why the wrong information may have been given on purpose.
A derivative record is created from a prior record. Therefore, all officially issued copies of records from an Italian town hall [whether an estratto, in international format, or a certificato] are derivative records. However, a photocopy of a civil record or an image copy, as found on FamilySearch microfilm, are not considered derivative as long as no changes were made to the record. The act of translation also makes a record derivative.
There are three main categories of genealogical information, primary, secondary, and indeterminable.
1. Primary information is reported by an eyewitness to the event, either at the time of the event or many years later. Therefore, is a death record recorded four days later by two men who lived on the same street as the deceased primary information? I would say for most of the information contained, yes. If they didn't directly see the person slip from life's grasp, then they likely saw the priest arrive to give the final sacraments, heard the weeping, or saw the civil official come to verify the death. That is primary information of the deceased's death date.
2. Secondary information is hearsay. Therefore, an Italian-American descendant who says their great-grandfather was killed in Palermo by the Mafia in 1904 is likely stating information they heard passed down through their family. Unless they have the death record stating his violent death in this city or some Italian court records documenting the murder, then it is likely hearsay.
3. Indeterminable information comes when we cannot determine who might have been the informant, therefore making it impossible to say with any accuracy what type of information it might be. In the death record example above, would a neighbor have primary information of the deceased's parent's names? Maybe or maybe not, making this information indeterminable or at the best, secondary, if it can be determined that they were told the information by someone else.
As in American documents, different pieces of information within a record could be different types of information. Cultural context and the practices behind the records also need to be considered.