Introduction

What follows is an attempt to describe the movement from an abstract data model and a semantically meaningful ontology to a serialization of that information that a generic web application can consume and display.

I will first provide a general overview of the abstract data model (visually represented here: https://raw.githubusercontent.com/scta/scta-ontology/master/SCTASCHEMA3.png) and how this model can be translated into a set of classes and subclasses that take on a logical and predictable set of properties suited to their class type (described herehttps://github.com/scta/scta-ontology/blob/master/SCTAOntologySpec.md)

Second, I will turn to a mock serialization of this same information that will illustrate how a web application could efficiently consume and display this information while minimizing HTTP requests.

In the end, I’ll demonstrate this entire movement from data model to serialization to web application consumption for a number of common resource types.

1. Data Model and Semantic Properties

A. WorkGroups and Works

The model begins from the concept of a top level WorkGroup. A top level WorkGroup can contain nested WorkGroups as well as a FRBR Work. I understand a FRBR Work to be an instance of the lowest level WorkGroup. This means that it has all the characteristics and properties of any other WorkGroup except that it is the lowest level and therefore cannot contain any other WorkGroups or Works. It can only contain FRBR Expressions.

Digression on FRBR CTS realignment

Right away, it should be noted that this is a departure from the way CTS (Canonical Text Services) has understood the term Work (see: http://www.homermultitext.org/hmt-docs/specifications/ctsurn/specification.html. While CTS has claimed an equivalence between a FRBR Work and a CTS Work, I have opted for a different approach.

CTS states:

Within a text group, a CTS URN’s work is a conceptual entity, like the FRBR work: it is an abstract idea of the content expressed in all versions of a work, in the original language or in translation. http://www.homermultitext.org/hmt-docs/specifications/ctsurn/specification.html

CTS moves directly from CTS Work to CTS Version which it identifies with a FRBR Expression, and from CTS Versions to CTS Exemplars which it identifies with FRBR Items. Missing from this, as CTS notes, is the concept of FRBR Manifestation.

Thus, we have the following alignment:

FRBR Work CTS Work
FRBR Expression CTS Version
FRBR Manifestation  
FRBR Item CTS Exemplar

This close, but not exact, alignment of the two models seems to be the source of great confusion, and I think we need to create a remapping and then revert to the FRBR language.

In brief, I think the deep abstractness of the FRBR Work is under appreciated. Likewise, our familiarity with the colloquial use of the word “work” to refer to an artistic creation that appears in a certain form (i.e. a book), inclines us to identify FRBR Work with what CTS means by Work. But in actual fact what CTS usually means by Work more closely parallels what FRBR means by Expression.

Below are the brief descriptions that FRBR gives for a FRBR Work and FRBR Expression:

Work is a “distinct intellectual or artistic creation.” For example, Beethoven’s Ninth Symphony apart from all ways of expressing it is a work. When we say, “Beethoven’s Ninth is magnificent!” we generally are referring to the work. Expression is “the specific intellectual or artistic form that a work takes each time it is ‘realized.’“[1] An expression of Beethoven’s Ninth might be each draft of the musical score he writes down (not the paper itself, but the music thereby expressed). https://en.wikipedia.org/wiki/Functional_Requirements_for_Bibliographic_Records

Critical in the definition of a FRBR Work is the phrase: considered “apart from all ways of expressing it.” Here we are very far from any notion of a text (a FRBR Expression) much less the idea of a typeset book (a FRBR Manifestation) and even further from the notion of a particular instance of a FRBR Manifestation in a library or bookshop somewhere in the world (a FRBR Item).

The idea of Moby Dick can be expressed in multiple ways. It can be expressed as a novel or as a screen-play. Each of these are separate FRBR Expressions that, as of yet, have made no reference to a particular edition or version. They are still just ideas. Each of these Expressions can then manifest themselves in different ways.

Thus from the outset I think we need a realignment of the two models that contradicts that alignment proposed by CTS.

Instead of:

FRBR Work CTS Work
FRBR Expression CTS Version
FRBR Manifestation  
FRBR Item CTS Exemplar

we should adopt the following alignment:

FRBR Work  
FRBR Expression CTS Work
FRBR Manifestation CTS Version
FRBR Item CTS Exemplar

I pause here to provide one last real world justification for this correction:

Adam Wodeham’s Oxford Ordinatio Commentary on the Sentences of Peter Lombard is a FRBR Work as I understand it.

But this Work was expressed by Wodeham in the 1330’s as his own written expression. For moment let us call this Adam Wodeham’s “Original Expression.”

But this same Work was expressed as an “Abbrevation” by Henry Totting of Oyta some 60 years later. Let’s call this the “Abbreviated Expression.” Henry Totting of Oyta took the same abstract Work and created a shorter abbreviated Expression, in a manner similar to the way a screen play might shorten or cut out sections left in a longer novel Expression. The screen play, like the “Abbreviated Expression” is a not a new CTS Version or FRBR Manifestation of the Work. It is a new Expression of the same idea that the “Original Expression” expresses.

In short, despite differences, the “Original Expression” and “Abbreviated Expression” are linked because they are two different Expressions of the same FRBR Work. Adam Wodeham is the author in both cases and they contain the same abstract content. But the content is expressed differently. The “Original Expression” was expressed by Wodeham himself while the “Abbreviated Expression” was expressed by Henry Totting of Oyta.

We cannot simply call these CTS Versions or FRBR Manifestations because each of these FRBR Expressions will have their own CTS Versions / FRBR Manifestations. For example, in 1516 John Mair produced the 1516 printed edition (FRBR Manifestation) of the “Abbreviated Expression”. This 1516 FRBR Manifestation was printed several times such that there are now several FRBR Items or CTS Exemplars of this particular Manifestation in different libraries throughout the world.

B. Expressions and OHCO

Once at the FRBR Expression, we have reached what we most commonly think of as the abstract text without yet specifying a particular material instantiation, version, or edition of that text. A good indicator that one has reached the FRBR Expression level is that the “text” can be broken down into a hierarchy. No such hierarchy is possible when discussing a FRBR Work because the idea has not been expressed in any form. A hierarchy only forms, when an author tries to express that idea. Accordingly, the hierarchies will likely differ in two different Expressions of the same Work.

Once at the Expression level, the existing text hierarchy makes it possible to construct a matrix, where FRBR concepts continue to run along the X axis and the hierarchy division runs down the Y axis. Thus, each unit within the FRBR Expression hierarchy can have FRBR Manifestations and FRBR Items.

While the hierarchy of a text is often unique to the Expression or to Expressions within a similar WorkGroup, it is possible to abstract from the details of the hierarchy and identify common patterns within any given hierarchy. Two obvious examples of this are the top level of the Expression hierarchy and the bottom level of the hierarchy. Every hierarchy, no matter how big or small, simple or complicated, will have a top level Expression and a bottom level Expression. Moreover, we can expect that these kinds of Expressions will have features or properties unique to their type. For example, a top level Expression will, by definition, never have a isPartOfTopLevelExpression property because there are no Expressions containing it. Likewise, a bottom level Expression will never have a hasPart property because it is the lowest level of the hierarchy and therefore there are no more parts to be contained.

Expression Structure Types and Expression Levels

The desire to identify general classes of resources within an Expression hierarchy while remaining agnostic about the actual hierarchy of any given Expression is the impetus behind the associated properties I refer to as structureTypes and levels.

The structureType property is used to divide a hierarchy, no matter its actual structure, into three privileged parts. A top level Expression (as we will see below, normally a top level structureCollection), a structureItem level, and structureBlock level.

Let’s consider first the structureItem division. This division, perhaps despite appearances, actually follows an established and traditional practice of dividing an Expression into a privileged set of units. In most books, this is the chapter division level. Consider the formatting of most monographs. Though there may be sections within books that contain chapters, and though chapters may contain many subdivisions, the chapter division is privileged. The chapter division gets a new page and a grand heading in a way that subdivisions do not. Likewise, if paragraph numbers, line numbers, or note numbers, get counted the numbering usually restarts at these privileged points in the Expression hierarchy. This seems to be clear evidence that this division within the hierarchy is being privileged. This privileging is not without good reason. It allows us, no matter how complicated the hierarchy of a text, to quickly jump from the top level Expression to exhaustively access all of the discrete sections within a text without having to loop through an unknown number of nested divisions.

The structureBlock performs a similar function. The structureBlock designation refers to the lowest exhaustive unit of text within the hierarchy. It is not possibly to identify the lowest unit in the Expression hierarchy simple by its level alone. Expression hierarchies may vary drastically in the number of levels within the hierarchy. So in order to easily identify the lowest exhaustive unit of a text (normally paragraphs or lines), we can designate each of these units as structureType=structureBlock.

Thus in two steps, no matter how complicated the hierarchy or how large the text, we can request every paragraph or every line in a text. We simply ask the top level Expression for every structureItem and every structureItem for every structureBlock.

Besides structureItem and structureBlock, we also use the notion of a structureCollection and structureDivision to classify other sections of the Expression hierarchy that are not otherwise designated.

A structureCollection is any section that stands above the structureItem level. These types help create predictable properties. Any Expression designated as a structureCollection would be expected to list all the structureItems it contains.

In this case, a top level Expression would also be a structureCollection. It would be differentiated from any other sub sections that are also structureCollections because it alone has a level property whose value equals “1”.

Any structureCollection whose level property does not equal 1 would be expect to also have a property isPartOfTopLevelExpression.

A structureDivision is similar. It represents a section in the hierarchy that is contained by a structureItem section, but is not a structureBlock section. Any structureDivision Expression should be expected to have the properties isPartOfStructureItem and hasStructureBlock.

C. Manifestations

FRBR Manifestations should now exist at every level of the Expression hierarchy. They can be associated through predictable properties. Each Expression can take a hasManifestation property and each Manifestation can take an isManifestationOf property.

Manifestations now bring us closer to the familiar CTS Version, but these versions vary in type. A Manifestation could be a manuscript, an early modern printing, a modern critical edition, or even a born digital critical edition.

It’s important to remember that even though a Manifestation is beginning to point to a physical or material representation of an Expression, it is nevertheless still an idea and not a physical object. We are referring here to the idea of the physical representation of an Expression. Thus when we talk about the 1512 printing of a given Expression, we are not talking about the Item of the 1512 printing in the British Library or in the Harvard University Library.

This is easier to see in the case of printed books, but harder to see in the case of a manuscript, which by definition will have only one material instantiation (or FRBR Item). Nevertheless, the conceptual distinction is important. The Manifestation that has the manifestationType manuscript will have one and only one material instantiation, whereas the Manifestation that has the manifestationType of a printed book will have several, if not thousands, of material instantiations.

Finally, we must also include a space here for born digital Manifestations of an Expression. The new practice of building born-digital critical editions of Expressions is a good example. This means that we have a Manifestation that corresponds to no physical material object. And it is of course tempting here associate this Manifestation with the actual digital file. But we must resist this temptation and continue to remember that the Manifestation is still an idea of how to materially represent (in this case, encode) a given Expression. It is the idea of a particular edition that may either have a material object or digital transcriptions standing as an instantiations of that idea.

In concrete, a Manifestation should inherit the structureType of the Expression it isManifestationOf. Moreover, a Manifestation should take on properties specific to a Manifestation. A Manifestation should have a manifestationType that indicates whether it is a manuscript, early printing, modern critical edition, or born digital critical edition. Other properties and constraints specific to each type of Manifestation could then be added.

Finally, a Manifestation should also point to FRBR Items. For reasons to be explained below we divide these associations into two types, hasMaterialObject and hasTranscription.

D. Items: MaterialObjects and Transcriptions

There seems to be something conceptually correct about identifying a digital transcription (by which I mean a semantically encoded text that can be serialized and distributed in multiple ways, and thus we do not yet mean the raw TEI file, or another other format) with a FRBR Item. Just like an individual printed books, it aims to be an instantiation of a Manifestation. However, instead of instantiating that idea with ink on a page, it does so with electronic signals on a disk drive.

Nevertheless, the digital transcription is also so different from the material instantiation that we felt it best to split the FRBR Item class into two subclasses: the MaterialObject and the Transcription.

The MaterialObject should be a record that has the appropriate properties to locate and describe the material book in this or that particular library.

My focus here will be on the Transcription subclass.

As mentioned above, the Transcription resource is not an actual digital file. Rather it is a resource that contains properties that can direct a client to the raw text representing this transcription in various file formats, the most obvious being XML. The hasXML property should point directly to the raw TEI file, which when requested through a traditional HTTP request returns exactly that: a TEI file.

Keep in mind that the the Transcription resource continues to inherit its hierarchical position from the Expression to which it is connected. It will rarely be the case then that separate source TEI XML files are produced for every Expression level. Rather, a Document or file will exist corresponding to some level of the Expression hierarchy. I find that it generally makes the most sense to divide my Documents at the structureItem level. This mean that the raw XML for a Transcription corresponding to structureCollection level will actually be composed from several documents. I expect that the hasXML property would point to a XML TEI document that is composed dynamically from the Documents required for its composition. Thus the hasDocument property should point to the XML Documents from which this Transcription’s raw XML TEI document was dynamically composed.

The same situation occurs for a structureDivision and structureBlock. The hasXML property should point to a raw XML file that is dynamically constructed from selecting a part of the Document in which the raw source XML is contained. In this case, there should only be one hasDocument property and it should point to the XML Document from which the raw XML for this Transcription was pulled and dynamically reconstructed.

The Transcription resource could also point to a number of other encoding serializations such as HTML or JSON or PlainText with parallel properties like hasHTML, hasJSON, or hasPlainText.

2. Serialization for a API consuming client

In this section, I’m going to propose some basic mock jsonld serializations of different resources. The expected properties and values returned in the serialization will reflect the data model described above and the nature of the resource in question. In connection with each serialization, I’ll provide a screen shot of a generic web application that could make use of this information.

Let’s start with a top level WorkGroup. A simple request for a top level WorkGroup would return the following jsonld data.

{
  "@id": "http://scta.info/resource/scta",
  "@type": "workGroup",
  "title": "Sentences Commentary Text Archive",
  "description": "The top level work group",
  "level": "1",
  "parts": [
    {
        "@id": "http://scta.info/resource/sententia",
        "@type": "workGroup",
        "description": "A Work Group for all commentaries on the Sentences of Peter Lombard",
        "title": "Sentences Work Group"
    },
    {
        "@id": "http://scta.info/resource/deanima",
        "@type": "workGroup",
        "description": "A Work Group for all commentaries on Aristotle's De Anima",
        "title": "De anima commentaries work group"
    }
  ],
  "expressions": [
    {
        "@id": "http://scta.info/resource/plaoulcommentary",
        "@type": "expression",
        "description": "Commentary on the Sentences by Peter Plaoul",
        "title": "Plaoul Commentary",
        "author": "Peter Plaoul"
    },
    {
        "@id": "http://scta.info/resource/wodehamordinatio",
        "@type": "expression",
        "description": "Commentary on the Sentences by Adam Wodeham",
        "title": "Wodeham Ordinatio",
        "author": "Adam Wodeham"
    },
    {
        "@id": "http://scta.info/resource/wodehamabbreviatio",
        "@type": "expression",
        "description": "Commentary on the Sentences by Adam Wodeham",
        "title": "Wodoeham Abbreviatio",
        "author": "Adam Wodeham"
    },
    {
        "@id": "http://scta.info/resource/rothwellcommentary",
        "@type": "expression",
        "description": "Commentary on the Sentences by William Rothwell",
        "title": "Rothwell Commentary",
        "author": "William de Rothwell"
    },
    {
        "@id": "http://scta.info/resource/meleandercommentary",
        "@type": "expression",
        "title": "Meleander de anima commentary",
        "author": "Placidus Aegidius Meleander"
    }
  ]
}

This can be displayed as follows:

top level work group

In the next serialization, we show a sample from the next (or second) level WorkGroup

{
  "@id": "http://scta.info/resources/sententia",
  "@type": "workGroup",
  "title": "Sententia",
  "description": "A work group of Sentences Commentaries",
  "level": "2",
  "isPartOf": "http://scta.info/resources/scta",
  "parts": [
    {
        "@id": "http://scta.info/resources/plaoulcommentary",
        "@type": "expression",
        "title": "Plaoul Commentary",
        "author": "Peter Plaoul"
    },
    {
        "@id": "http://scta.info/resources/wodehamordinatio",
        "@type": "expression",
        "title": "Wodeham Ordinatio",
        "author": "Adam Wodeham"
    },
    {
        "@id": "http://scta.info/resources/wodehamabbreviatio",
        "@type": "expression",
        "title": "Wodoeham Abbreviatio",
        "author": "Adam Wodeham"
    },
    {
        "@id": "http://scta.info/resources/rothwellcommentary",
        "@type": "expression",
        "title": "Rothwell Commentary",
        "author": "William de Rothwell"
    }
    ],
  "expressions": [
    {
        "@id": "http://scta.info/resources/plaoulcommentary",
        "@type": "expression",
        "title": "Plaoul Commentary",
        "author": "Peter Plaoul"
    },
    {
        "@id": "http://scta.info/resourcees/wodehamordinatio",
        "@type": "expression",
        "title": "Wodeham Ordinatio",
        "author": "Adam Wodeham"
    },
    {
        "@id": "http://scta.info/resources/wodehamabbreviatio",
        "@type": "expression",
        "title": "Wodoeham Abbreviatio",
        "author": "Adam Wodeham"
    },
    {
        "@id": "http://scta.info/resources/rothwellcommentary",
        "@type": "expression",
        "title": "Rothwell Commentary",
        "author": "William de Rothwell"
    }
  ]
}

The following image, shows a rendering of the above WorkGroup, where the client has chosen to show all the Expressions contained by a WorkGroup.

top level work group

The next serialization shows the jsonld response for a top level Expression:

{
  "@id": "http://scta.info/resource/plaoulcommentary",
  "@type": "expression",
  "title": "Plaoulcommentary",
  "author": "Peter Plaoul",
  "description": "Commentary on the Sentences by Peter Plaoul",
  "structureType": "collection",
  "level": "1",
  "isPartOf": "http://scta.info/resource/sententia",
  "canonicalManifestation":
    {
        "@id": "http://scta.info/resource/plaoulcommentary/critical",
        "title": "Critical Edition for Plaoul Commentary",
        "type": "critical",
        "status": "Draft",
        "editor": "Jeffrey C. Witt",
        "canonicalTranscription":
        {
            "@id": "http://scta.info/resource/plaoulcommentary/critical/transcription",
            "title": "Critical Edition Transcription for Plaoul Commentary",
            "type": "critical",
            "status": "Draft",
            "editor": "Jeffrey C. Witt",
            "xml": "http://scta.info/text/plaoulcommentary.xml"
        }
    },
    "manifestations": [
    {
        "@id": "http://scta.info/resource/plaoulcommentary/reims",
        "title": "Reims Manuscript for Plaoul Commentary",
        "type": "manuscript"
    },
    {
        "@id": "http://scta.info/resource/plaoulcommentary/vat",
        "title": "Vatican Manuscript for Plaoul Commentary",
        "type": "manuscript"
    },
    {
        "@id": "http://scta.info/resource/plaoulcommentary/svict",
        "title": "St. Victor Manuscript for Plaoul Commentary",
        "type": "manuscript"
    },
    {
        "@id": "http://scta.info/resource/plaoulcommentary/sorb",
        "title": "Sorbonne Manuscript for Plaoul Commentary",
        "type": "manuscript"
    },
    {
        "@id": "http://scta.info/resource/plaoulcommentary/critical",
        "title": "Critical Edition for Plaoul Commentary",
        "type": "critical"
    }
  ],
  "parts": [
    {
        "@id": "http://scta.info/resource/deFide",
        "title": "De Fide"
    },
    {
        "@id": "http://scta.info/resource/deNotitia",
        "title": "De Notitia"
    },
    {
        "@id": "http://scta.info/resource/deFruitione",
        "title": "De Fruitione"
    },
    {
        "@id": "http://scta.info/resource/deTrinitate",
        "title": "De Trinitate"
    },
    {
        "@id": "http://scta.info/resource/deCaritate",
        "title": "De Caritate"
    },
    {
        "@id": "http://scta.info/resource/deCaritate",
        "title": "De Caritate"
    },
    {
        "@id": "http://scta.info/resource/deLibertate",
        "title": "De Libertate"
    },
    {
        "@id": "http://scta.info/resource/deIncarnatione",
        "title": "De Incarnatione"
    }
  ],
  "items": [
    {
        "@id": "http://scta.info/resource/lectio1",
        "title": "Lectio 1",
        "status": "Draft",
        "gitRepo": "http://bitbucket.org/jeffreycwitt/lectio1"
    },
    {
        "@id": "http://scta.info/resource/lectio2",
        "title": "Lectio 2",
        "status": "Draft",
        "gitRepo": "http://bitbucket.org/jeffreycwitt/lectio2"
    },
    {
        "@id": "http://scta.info/resource/lectio3",
        "title": "Lectio 3",
        "status": "Draft",
        "gitRepo": "http://bitbucket.org/jeffreycwitt/lectio3"
    },
    {
        "@id": "http://scta.info/resource/lectio4",
        "title": "Lectio 4",
        "status": "Draft",
        "gitRepo": "http://bitbucket.org/jeffreycwitt/lectio4"
    },
    {
        "@id": "http://scta.info/resource/lectio5",
        "title": "Lectio 5",
        "status": "Draft",
        "gitRepo": "http://bitbucket.org/jeffreycwitt/lectio5"
    },
    {
        "@id": "http://scta.info/resource/lectio6",
        "title": "Lectio 6",
        "status": "Draft",
        "gitRepo": "http://bitbucket.org/jeffreycwitt/lectio6"
    },
    {
        "@id": "http://scta.info/resource/lectio7",
        "title": "Lectio 7",
        "status": "Draft",
        "gitRepo": "http://bitbucket.org/jeffreycwitt/lectio7"
    },
    {
        "@id": "http://scta.info/resource/lectio8",
        "title": "Lectio 8",
        "status": "Draft",
        "gitRepo": "http://bitbucket.org/jeffreycwitt/lectio8"
    }
  ]
}

Here’s how a client might consume that information to provide information about nested parts:

top level work group

Finally, we can illustrate what the jsonld might look like for a non top-level Expression, such as an Expression with a structureType=structureItem

{
  "@id": "http://scta.info/resource/lectio1",
  "@type": "expression",
  "title": "Lectio 1",
  "author": "Peter Plaoul",
  "description": "Lectio 1 from Plaoul's Commentary",
  "structureType": "item",
  "level": "3",
  "isPartOf": "http://scta.info/resource/deFide",
  "topLevelExpression": "http://scta.info/resource/plaoulcommentary",
  "next": "http://scta.info/resource/lectio2",
  "canonicalManifestation":
    {
        "@id": "http://scta.info/resource/lectio1/critical",
        "title": "Critical edition for lectio 1 in Plaoul Commentary",
        "type": "critical",
        "status": "Draft",
        "editor": "Jeffrey C. Witt",
        "canonicalTranscription":
        {
            "@id": "http://scta.info/resource/lectio1/critical/transcription",
            "title": "Critical Edition Transcription for lectio 1 in Plaoul Commentary",
            "status": "Draft",
            "editor": "Jeffrey C. Witt",
            "xml": "http://bitbucket.org/jeffreycwitt/raw/master/lectio1.xml"
        }
    },
  "manifestations": [
    {
        "@id": "http://scta.info/resource/lectio1/reims",
        "title": "Reims Manuscript for lectio 1 in Plaoul Commentary",
        "type": "manuscript"
    },
    {
        "@id": "http://scta.info/resource/lectio1/vat",
        "title": "Vatican Manuscript for lectio 1 in Plaoul Commentary",
        "type": "manuscript"
    },
    {
        "@id": "http://scta.info/resource/lectio1/svict",
        "title": "St. Victor Manuscript for lectio 1 in Plaoul Commentary",
        "type": "manuscript"
    },
    {
        "@id": "http://scta.info/resource/lectio1/sorb",
        "title": "Sorbonne Manuscript ffor lectio 1 in Plaoul Commentary",
        "type": "manuscript"
    },
    {
        "@id": "http://scta.info/resource/lectio1/critical",
        "title": "Critical Edition for lectio 1 in Plaoul Commentary",
        "type": "critical"
    }
  ],
  "parts": [],
  "blocks": [
    {
        "@id": "http://scta.info/resource/p1",
        "title": "paragraph 1"
    },
    {
        "@id": "http://scta.info/resource/p2",
        "title": "paragraph 2"
    },
    {
        "@id": "http://scta.info/resource/p3",
        "title": "Paragraph 3"
    },
    {
        "@id": "http://scta.info/resource/p4",
        "title": "Paragraph 4"
    }
  ]
}

And this information can be used as follows:

item text view

But this same information can be used for a different view, such as a view of available Manifestations:

available manifestations

Likewise, the client should be able to treat every Expression, at any level in the hierarchy, the same way while adjusting for properties unique to that structureType. So if the resource id points to a structureBlock (or paragraph) instead of a structureItem, we get a display that automatically adjusts.

block level display

3. Appendices

Appendix A: Data Model Visualization

https://raw.githubusercontent.com/scta/scta-ontology/master/SCTASCHEMA3.png

https://raw.githubusercontent.com/scta/scta-ontology/master/SCTASCHEMA3.png

Appendix B: Resource Types and Properties List

https://github.com/scta/scta-ontology/blob/master/SCTAOntologySpec.md