Typo indexes are confusing

[joeroe]

Many sites will have a long list of identical typological definitions repeated many times (one for each radiocarbon date).

Technically this is correct, according to our data model: each typo represents a determination of an individual sample (the one taken for radiocarbon dating) according to our source database. However, this is not apparent to the user and probably just looks messy.

I can see two options:

1. Stick to the strict interpretation of the data model and improve the view to make it clearer that each typo applies to a different sample, perhaps by generating unique labels for samples where they are not present
2. The less principled but cleaner option – ditch the duplicates and ensure the there is just one occurrence of each typo determination per site, probably attached to a new dummy sample rather than the radiocarbon samples

[MartinHinz]

I guess, you also feel that option two would be better? How about counting the mentioning of a specific typo instead of listing all of them?

[joeroe]

Actually I feel better about option 1, but I can see there's a gap between the data-modelling theory (if someone puts "Neolithic" in a row next to a radiocarbon date, they're implying that sample specifically is Neolithic) versus the reality (they probably didn't mean it that way). I can't make my mind up.

A middle-ground option 3: aggregate typos by context, rather than site? So we can have more than one "Neolithic" attached to Site X, but only if they are associated with different contexts?