"Basket" and "Lot" are types of sub-site contexts / locations where stuff was found. Do they belong in anthropogenicAggregation?

[dannymandel]

"Basket" and "Lot" are types of sub-site contexts / locations where stuff was found.

_Originally posted by @ekansa in https://github.com/isamplesorg/metadata/pull/51#discussion_r667241617_

[dannymandel]

@smrgeoinfo Could you confirm "Basket" and "Lot" belong in antropogenicAggregation or suggest an alternative?

[smrgeoinfo]

@ekansa I'm a little mystified by the 'basket'. Spot checking the baskets at https://opencontext.org/subjects-search/?prop=oc-gen-cat-basket, they each seem to be a collection of artifacts (or bones...), many of which are associated with a stratum (e.g. 9th c, 9th/8th, Late Bronze) or phase (e.g. II, street), and the baskets all seem to be associated with a Locus. Seems like the 'Locus' is the most specific site information. Is the idea that the stuff in each basket was collected from the same part of the Locus? Since most of the basket contents I looked at seem to be bones, 'Anthropogenic aggregation' is based on the idea is that the bones are the result of human activity (animal butchering, meat consumption); if not then 'Aggregation' would be more appropriate. Not a great fit either way.

[smrgeoinfo]

solution might be 'basket' is in the location information, based loosely on https://opencontext.org/subjects/HazorZooSPA0000012004, I'm not sure if the information can be mined from the raw JSON we're working with:

     "description": "set of bones collected from a particular spot (?pit?) in Locus 5794 at Hazor excavation site",
    "hasContextCategory": ["Site of past human activities"],
    "hasMaterialCategory": ["Biogenic non-organic material"], (assuming its a basket of bones)
    "hasSpecimenCategory": ["Anthropogenic aggregation"], (or 'Aggregation' if the bones are not the result of human activity)
    "producedBy": {
        "label": "sampling event",
        "description": "",
        "hasFeatureOfInterest": "Tel Khatsor",
        "responsibility": ["Justin Lev-Tov"],
        "resultTime": "",
        "samplingSite": {
            "description": "Israel, Hazor,  Area M,  Locus 5794, Basket 52065",
            "label": "collection of items from a particular part of locus in excavation site; place names should be at the level of the reported lat/long, see table at http://wiki.gis.com/wiki/index.php/Decimal_degrees. 6 decimal places corresponds to about 10 cm precision",
            "location": {
                "elevation": "",
                "latitude": 33.01, 
                "longitude": 35.56
            },
            "placeName": ["Israel", "Hazor", "Area M", "Tel Khatsor", "Tel Hazor National Park"]
        }
    }

[smrgeoinfo]

Looks like 'lot' is pretty much like basket, but there seem to be detail maps of the excavation site showing the position from which the 'lot' was collected. The content of the lot (e.g. https://opencontext.org/subjects/1217_DT_Spatial) seems to be sets of bones or artifacts.

[ekansa]

@smrgeoinfo there's some "strategic ambiguity" here. For background, archaeologists have different recording systems with somewhat different notions of stratigraphic / contextual units.

• A "lot" in Domuztepe is like a "Locus" from other projects (like Poggio Civitate / Murlo). It's a minimal stratigraphic / contextual unit that's surmised to be the result of some given process of deposition (filling pit with garbage, accumulation of debris from the deterioration of a mudbrick structure, construction of a wall, etc)

A "basket" is usually an arbitrarily defined part of a given stratigraphic / contextual unit. Baskets typically get used to help provide some protection against mixing finds from different stratigraphic units. During excavations, archaeologists may miss interfaces / transitions between different stratigraphic units (sometimes these are subtle, sometimes the excavators lack experience, etc). Sometimes you only notice a transition late. If you use baskets to sub-divide material from a stratigraphic / context unit, then the baskets could limit the damage of mixing finds from different units. You can say, these baskets are firmly in Locus 1, this basket mixes locus 1 and 2, and these other baskets are firmly in Locus 2.

Open Context has very loose semantics for this type of thing. We basically just model a basket as a child context of some parent context. I'd expect some variation in how archaeologists think about how a "basket" works and some other archaeologists may also record "baskets" but not really think of them as ways to protect against mixing materials. I believe some archaeologists may use baskets to group pottery, lithics, or animal bones from a given stratigraphic unit.

In any event, in the iSamples scheme, I guess a basket can typically be a "sampling event" of a "sampled feature" (a stratigraphic unit / context unit)? Am I understanding the iSamples organization OK?

[smrgeoinfo]

The question is about specimenType for the content of the basket or lot (assuming that content is considered the physical sample). Sounds like the content is a 'bunch of items found at a particular place', the specimen type would be 'Aggregation', or if all the bits of stuff in the basket/lot are known to be artifacts, then it would be 'Anthropogenic Aggregation'.

Can we close this one?