Primary skeletal inventories
Primary skeletal inventories are skeletal inventories that are exclusively used for entering skeletal material into the database for the first time. They establish the material basis for all investigations carried out on the material kept in the skeletal collection that is managed by the database. All other skeletal inventories are limited to referencing skeletal elements that are part of some primary skeletal inventory. This is to ensure that each skeletal element in the collection is represented by exactly one instance in the database and to prevent skeletal elements to be entered repeatedly as different entities by different users.
As primary skeletal inventories are exclusively concerned with establishing the existence of skeletal material, they only record their presence, not their condition. As a consequence, primary skeletal inventories do not contain measurement data of type rdfbones:TaphonomicStateDatum (cf. the page on skeletal inventories. To ensure that skeletal material is registered consitently throughout the database, primary skeletal inventories have a very simple structure that is defined in the RDFBones core ontology and cannot be altered through ontology extensions.
Primary skeletal inventories are compiled on the basis of individual bone organs and teeth, as they represent natural entities that are unambiguously identified. As a consequence, it is necessary to discern if a skeletal element is present in its entirety or just in parts, by means of measurement data of type rdfbones:CompletenessDatum (cf. the page on skeletal inventories. With incomplete skeletal elements, there is a danger of registering various fragments of the same element as separate instances. This is particularly relevant for commingled assemblages. If a skeletal element is accidentaly entered twice, identity of the two instances has to be established through the object property owl:sameAs (cf. OWL Web Ontology Language Reference). Registering fragments instead of skeletal elements is not a viable alternative, as they are not natural entities and their number might increase any time.
Skeletal elements that are firmly joined together clearly belong to the same skeleton. Therefore, it makes a difference if two parietal bones are found separately or if they are fused at the saggital suture to form a skull. This type of information is captured by primary skeletal inventories. If skeletal elements form a superordinate skeletal structure, this structure is also recorded in the inventory (represented by an instance of obo:FMA_53672 'Viscerocranium' in the example), along with the subordinate elements (instances of _obo:FMA\9613 'Parietal bone') that are part of it (obo-fma:systemic_part_of). This is not the case if skeletal elements are found separately, even if a researcher might suspect skeletal elements to belong to the same skelelton.
Figure: Measurement data employed in primary skeletal inventories and their application (cf. key to symbols in network graphs).
Primary skeletal inventories contain a limited set of measurement data to record presence and completeness of skeletal features.
Measurement data of type rdfbones:PhenotypicExpressionDatum are used to express if a biological trait expected in a species is actually realised in an individual (cf. the page on skeletal inventories). With primary skeletal inventories, the sole interest is whether a trait is expressed or not, requiring two states of phenotypic expression: 'expressed' and 'not expressed'. These are the categorical labels that are available for measurement data of type rdfbones:PhenotypicExpression2States, represented by a subclass of rdfbones:PhenotypicExpressionDatum.
In primary skeletal inventories, rdfbones:PhenotypicExpression2States is used to record if all tooth positions expected in a grown-up individual are realised (which is not true in cases of hypodontia). It has to be noted that the expected number of tooth positions is derived from the fact that at least one jaw bone is present, not from the circumstances under which the material recorded in the primary skeletal inventory was obtained. In this cicumstance, coding of missing skeletal features is needed to discern cases where a certain feature is not preserved from cases where it never existed.
Measurement data of type rdfbones:ObliterationDatum are used to express if a biological trait has been obscured or destroyed by some process (cf. the page on skeletal inventories). Such an occurrence would result in the fact that a biological feature is not present, but the implication would be different from a situation where the feature in question would have never existed (cf. 'Phenotypic expression in two states' above). In the simplest form, such a datum would require two levels, 'obliterated' and 'not obliterated'. These are the two available lables for measurement data of type rdfbones:Obliteration2States, represented by a subclass of rdfbones:ObliterationDatum.
Measurement data of type rdfbones:CompletenessDatum are used to express what portion of an element is present. Primary skeletal inventories use a subclass of rdfbones:CompletenessDatum, rdfbones:Completeness2States, to record whether skeletal elements are either complete or only partly present. This basic distinction is important, as entries for complete elements ensure unambiguous references of this element. Elements that come in several fragments, on the other hand, may be recorded repeatedly if their association to the same skeletal element is unknown at the time of data entry. This is particularly relevant for comingled remains.
There are several types of primary skeletal inventories defined in the RDFBones core ontology for entering different kinds of skeletal material.
Figure: General structure of the primary skeletal inventory for fresh bone (cf. key to symbols in network graphs).
Class PrimarySkeletalInventoryFreshBone represents an inventory for skeletal material that still contains a certain amount of its organic component (mainly collagen). Individual bones (obo:FMA_5018 'Bone organ') are reresented by subclasses of rdfbones:EntireBoneOrgan, all of them being skeletal segements (rdfbones:SegmentOfSkeletalElement) for which the data property rdfbones:skeletalElementPortion has a value of 1 (cf. the page on skeletal inventories).
Each subclass of rdfbones:EntireBoneOrgan has a matching subclass of rdfbones:Completeness2StatesEntireBoneOrgan, a subclass of rdfbones:Completeness2States (see above). These measurement data are used to record if the bone organs being entered are complete or only partially present.
Figure: General structure of the primary skeletal inventory for dry bone (cf. key to symbols in network graphs).
Analogous to primary inventories for fresh bone (see above), class rdfbones:PrimaryInventoryDryBone defines a primary skeletal inventory for human remains entirely depleted of their organic component. Typcially, this is the case with skeletal remains from archaeological contexts. Technically speaking, these bones are not preserved in their entirety, but only the mineral component, represented by subclasses of obo:FMA_83129 'Bony part of bone organ'. To account for this difference, the classes rdfbones:EntireBonyPartOfBoneOrgan and rdfbones:Completeness2StatesEntireBonyPartOfBoneOrgan are used here to take the funcitons of their counterparts in the skeeltal inventory for fresh bone.
Figure: General structure of the primary dental inventory (cf. key to symbols in network graphs).
Teeth have a primary skeletal inventory of their own, defined with class rdfbones:PrimaryDentalInventory, because they represent a particular situation. Teeth can be either found separately or embedded in a jaw bone. The presence of a jaw bone with empty tooth sockets is evidence for the fact that the teeth that used to sit in these sockets existed. Also, if a tooth socket in a jaw bone was closed during an individual's lifetime, this is evidence for the corresponding tooth being lost premortem and therefore its presence today is not to be expected. Hypodontia (dentitions with less teeth than the expected 32) can be evidenced by a limited number of tooth sockets, showing that teeth are not missing but never existed. These inferred information are recorded in primary dental inventories.
The presence and completeness of teeth is recorded similarly to the completeness data in the primary skeletal inventories for fresh and dry bone, with rdfbones:EntireTooth and rdfbones:Completeness2StatesEntireTooth replacing the corresponding classes there.
Tooth sockets are represented by skeletal segments of type rdfbones:EntireToothSocket. For each tooth socket observed in a jaw bone, it is recorded if it is closed as a consequence of antemortem tooth loos ('obliterated') or not ('not obliterated'), using measurement data of type rdfbones:Closure2StatesToothSockets, a subclass of rdfbones:Obliteration2States (see above).
Teeth, no matter if complete or only partly present, can be either found separately or in a tooth socket that, consequently is not obliterated. If teeth are encountered firmly rooted in a tooth socket, this fact is recorded by linking them to the respective socket through the object property obo:SIO_000128 ('is conatained in'). This is not applicable to teeth that fit well into the corresponding socket, so that their association with this socket is very likely. Primary inventories only record information that is absolutely conclusive and not based on any interpetation. The association of loose teeth with corresponding tooth sockets, therefore, has to be established with some ordinary skeletal inventory.
If a tooth socket is observed, this proves that the corresponding tooth position exists in the dentition.
Primary skeletal inventories (:PrimarySkeletalInventory) are referenced by a centrally registered identifier (obo:IAO_0000577 'Centrally registered identifier') that denotes a registry (obo:IAO_0000579 'centrally registered identifier registry') which acts as primary registry (:PrimaryRegistry) for a collection (cf. Identification and referencing systems). They are composed of categorical data items (:Completeness2State) that can have either of the two categorical labels 'complete' and 'partly present' (cf. How data are represented in OBI). All data refer to subclasses of either :EntireBoneOrgan or :EntireBonyPartOfBoneOrgan or their subclasses, implying the values for :bonePortion and :bonyPartOfBonePortion, respectively, to be 1.0.
The following graph illustrates classes and instances in a simple primary skeletal inventory. The rectangle contains data instances (edged boxes), outside the rectangle the containing classes (rounded edges) and their relations are depicted. The example contains two bones, a frontal bone and a right parietal bone, from a batch of skeletal material, referenced by the 'Material-ID' identifier 0815. The frontal bone is complete, the right parietal bone only partially present.
