TheSu XML (Thesis-Support XML) is a stand-off annotation schema for modelling ideas conveyed by textual sources and linking those ideas to their surrounding discourse contexts. The schema uses external pointers to reference text spans without embedding markup in source documents, preserving source integrity and structure and enabling compatibility with various source formats (TEI-encoded XML, XHTML, and others). Bibliographic entries use TEI encoding. The schema enables researchers to annotate declarative statements (theses) with thematic classifiers, formal structure, and speaker attribution; connect theses to discursive contexts (supports) that provide argumentative, expository, expansive, or contextualizing functions; and link multiple theses expressing similar core ideas under abstract propositions, facilitating systematic comparative analysis across texts and corpora.

Schema Reference

For comprehensive schema documentation, detailed attribute references, and relationship type specifications, consult the official documentation sources:

Complete Documentation

Full element and attribute reference with detailed specifications and examples.

View Documentation

Schema Definition

Official XSD schema file for validation and reference.

View XSD

⚠️ Why Are So Many Elements "Required"?

In the current version of the XSD, many elements or attributes are marked as "required". This is not always a semantic requirement—the schema is planned to become more flexible in the future. Currently, these requirements exist because TheSu XML is used within Oxygen XML, taking advantage of the "Author" mode automations together with CSS styling. These requirements speed up the annotation process by ensuring consistent structure.

When the dedicated TheSu Annotator GUI is released, most elements and attributes will become "optional", allowing for more flexible annotation workflows.

Note: Most examples throughout this documentation come from ancient Greco-Roman sources on lead and lead white, which form the case study for the FWO-funded project developing TheSu XML.

Main Elements

THESIS

Declarative statements that represent explicit or implicit claims within a source. Theses are linked to specific text spans and can be connected to supporting contexts and abstract propositions.

<THESIS xml:id="...">
  <speakersGroup>...</speakersGroup>
  <thesisType>...</thesisType>
  <text>...</text>
  <paraphrasis>...</paraphrasis>
</THESIS>

Child Elements

  • speakersGroup — Identifies the individuals or groups presented as speakers of the thesis. Contains one or more speaker elements, each with optional @rank attribute to indicate likelihood when identification is ambiguous.
  • thesisType — Annotates thematic classification and formal features. Contains macroThemesGroup for broad themes (e.g., "historical", "axiological"), microThemesGroup for detailed themes (e.g., "military", "psychological"), keywordsGroup for annotated keywords, definition if the thesis has definitional structure, sequencesGroup for processual analyses (recipes, procedures, chains of events), etiologiesGroup for causal structures, analogiesGroup for analogical relationships, and metaphorsGroup for metaphorical expressions.
  • text — References one or more segments of the primary source text via textRef elements, each containing segment elements pointing to text spans. May include differences for variant readings and uncertainties for corrupt or uncertain text.
  • paraphrasis — Provides a clear, concise interpretive paraphrase of the thesis's semantic content in the corpus language. Required for implicit or extrinsic theses, where the referenced text differs from the annotated meaning. Must be consistent with all other annotated details (polarity, thematic classification, formal structure, etc.).

This example shows a complete THESIS element from an annotated text. It includes a speaker attribution, thematic classification, a reference to the source text with the actual snippet, and a paraphrase.

<THESIS xml:id="tlg0059_tlg020.T181551">
  <matchingPropositionsGroup/>
  <speakersGroup>
    <speaker name="./meta/ancient-lead-white_JOHD_people.xml#Socrates"/>
  </speakersGroup>
  <assent>
    <assentSpeaker>
      <limitationsGroup/>
    </assentSpeaker>
    <assentSupSpeakersGroup/>
    <assentAuthor/>
  </assent>
  <thesisType>
    <macroThemesGroup>
      <macroTheme macroThemeTag="./meta/ancient-lead-white_JOHD_indexing-tags.xml#physical"/>
    </macroThemesGroup>
    <microThemesGroup>
      <microTheme microThemeTag="./meta/ancient-lead-white_JOHD_indexing-tags.xml#metaphysics"/>
    </microThemesGroup>
    <keywordsGroup>
      <keywordRef/>
    </keywordsGroup>
    <definition/>
    <sequencesGroup/>
    <etiologiesGroup/>
    <analogiesGroup/>
    <metaphorsGroup/>
  </thesisType>
  <text>
    <textRef>
      <segment from="./sources-segmented/tlg0059.tlg020.perseus-grc2.xml#w1004843" to="./sources-segmented/tlg0059.tlg020.perseus-grc2.xml#w1004856"/>
      <snippet>ἔνια μέν , οἷον ἂν ᾖ τὸ παρόν , τοιαῦτά ἐστι καὶ αὐτά , ἔνια δὲ οὔ</snippet>
    </textRef>
  </text>
  <paraphrasis>Some things are also themselves of such a sort as whatever may be present (in/to them); others are not</paraphrasis>
</THESIS>

Referenced segment in the segmented source file:

<p>
  <w xml:id="w1004840">λέγω</w>
  <space xml:id="sp104707"/>
  <w xml:id="w1004841">γὰρ</w>
  <space xml:id="sp104708"/>
  <w xml:id="w1004842">ὅτι</w>
  <space xml:id="sp104709"/>
  <w xml:id="w1004843">ἔνια</w>
  <space xml:id="sp104710"/>
  <w xml:id="w1004844">μέν</w>
  <pc xml:id="pc101230">,</pc>
  <space xml:id="sp104711"/>
  <w xml:id="w1004845">οἷον</w>
  <space xml:id="sp104712"/>
  <w xml:id="w1004846">ἂν</w>
  <space xml:id="sp104713"/>
  <w xml:id="w1004847">ᾖ</w>
  <space xml:id="sp104714"/>
  <w xml:id="w1004848">τὸ</w>
  <space xml:id="sp104715"/>
  <w xml:id="w1004849">παρόν</w>
  <pc xml:id="pc101231">,</pc>
  <space xml:id="sp104716"/>
  <w xml:id="w1004850">τοιαῦτά</w>
  <space xml:id="sp104717"/>
  <w xml:id="w1004851">ἐστι</w>
  <space xml:id="sp104718"/>
  <w xml:id="w1004852">καὶ</w>
  <space xml:id="sp104719"/>
  <w xml:id="w1004853">αὐτά</w>
  <pc xml:id="pc101232">,</pc>
  <space xml:id="sp104720"/>
  <w xml:id="w1004854">ἔνια</w>
  <space xml:id="sp104721"/>
  <w xml:id="w1004855">δὲ</w>
  <space xml:id="sp104722"/>
  <w xml:id="w1004856">οὔ</w>
  <pc xml:id="pc101233">.</pc>
</p>

Below is a detailed example showing the complete structure of thesisType, which contains all thematic and formal annotation for a thesis, including detailed examples of keywordsGroup and etiologiesGroup:

<thesisType>
  <macroThemesGroup>
    <macroTheme macroThemeTag="./meta/ancient-lead-white_JOHD_indexing-tags.xml#physical"/>
  </macroThemesGroup>
  <microThemesGroup>
    <microTheme microThemeTag="./meta/ancient-lead-white_JOHD_indexing-tags.xml#chemistry"/>
    <microTheme microThemeTag="./meta/ancient-lead-white_JOHD_indexing-tags.xml#cosmetics"/>
  </microThemesGroup>
  <keywordsGroup>
    <keywordRef ref="#keyword_id_1" focus="1"/>
    <keywordRef ref="#keyword_id_2" focus="2"/>
    <keywordRef ref="#keyword_id_3" focus="1"/>
  </keywordsGroup>
  <definition>...</definition>
  <sequencesGroup>...</sequencesGroup>
  <etiologiesGroup>
    <etiology>
      <marker>
        <textRef>
          <segment from="..." to="..."/>
        </textRef>
      </marker>
      <etiologyMember>
        <textRef>
          <segment from="..." to="..."/>
        </textRef>
      </etiologyMember>
      <etiologyMember cause="true">
        <textRef>...</textRef>
      </etiologyMember>
    </etiology>
  </etiologiesGroup>
  <analogiesGroup>...</analogiesGroup>
  <metaphorsGroup>...</metaphorsGroup>
</thesisType>

The thesisType element contains all the information about what a thesis is about and how it's structured.

Themes: macroThemesGroup and microThemesGroup classify the thesis by topic. Think of macro-themes as broad categories (like "physical" or "historical") and micro-themes as more specific topics (like "chemistry" or "cosmetics"). Theme tags always reference authority records (e.g., ./meta/ancient-lead-white_JOHD_indexing-tags.xml#physical) rather than using simple strings, ensuring consistency and standardization across the corpus.

Keywords: keywordsGroup lists the important words in the thesis. Each keywordRef points to a keyword that's been annotated elsewhere in the document. The @focus attribute shows how central each keyword is: "1" means most important, "5" means least important.

Keywords are annotated separately in the document so that multiple theses can reuse the same keyword. This is useful when several theses share the same important words—for example, when they discuss the same topic in the same passage, or when they reference the same concepts (even if some are implicit). Instead of annotating the same keyword multiple times, you annotate it once and then reference it from each thesis that uses it.

<keyword xml:id="tlg0059_tlg020.K182556" namely="./meta/ancient-lead-white_JOHD_indexing-tags.xml#colour">
  <text>
    <textRef>
      <segment from="./sources-segmented/tlg0059.tlg020.perseus-grc2.xml#w1004861"/>
    </textRef>
  </text>
  <microThemesGroup>
    <microTheme microThemeTag="./meta/ancient-lead-white_JOHD_indexing-tags.xml#chemistry"/>
    <microTheme microThemeTag="./meta/ancient-lead-white_JOHD_indexing-tags.xml#chromatology"/>
  </microThemesGroup>
</keyword>

Each keyword has its own unique ID (in @xml:id), points to the text where it appears, and can include thematic tags. The keywordRef elements in keywordsGroup reference these keywords using their IDs. In this example, the keyword "colour" is annotated with two micro-themes: "chemistry" and "chromatology", indicating that it relates to both chemical and color-related concepts.

Etiologies: etiologiesGroup captures statements about causes and effects. Each etiology represents one cause-and-effect relationship. It can include a marker showing words like "because" that signal the relationship. Each etiology has at least two etiologyMember elements—one for the cause (marked with @cause="true") and one for the effect (marked with @end="true"). Each member can reference its text in different ways, such as pointing to specific text segments or providing a paraphrase.

This example shows a complete etiology element from an annotated text. It demonstrates how cause-and-effect relationships are captured, with one etiologyMember representing the event and another marked with @cause="true" to identify the cause.

<etiologiesGroup>
  <etiology>
    <etiologyMember>
      <textRef>
        <segment from="./sources-segmented/tlg0007.tlg112.perseus-grc2.xml#w1043129" to="./sources-segmented/tlg0007.tlg112.perseus-grc2.xml#w1043130"/>
        <segment from="./sources-segmented/tlg0007.tlg112.perseus-grc2.xml#w1043137" to="./sources-segmented/tlg0007.tlg112.perseus-grc2.xml#w1043140"/>
        <snippet>τῶν ὑδάτων […] ὑπὸ τοῦ ψυχροῦ · κρατεῖται</snippet>
      </textRef>
    </etiologyMember>
    <etiologyMember cause="true">
      <textRef>
        <segment from="./sources-segmented/tlg0007.tlg112.perseus-grc2.xml#w1043142" to="./sources-segmented/tlg0007.tlg112.perseus-grc2.xml#w1043143"/>
        <snippet>διʼ ἀσθένειαν</snippet>
      </textRef>
    </etiologyMember>
  </etiology>
</etiologiesGroup>

In this example, the first etiologyMember (without attributes) represents the event: "thinner waters are dominated by the cold". The second etiologyMember has @cause="true", marking it as the cause: "because of their weakness". When an etiologyMember has no @cause or @end attribute, it represents an event whose cause or end is specified by sibling members. This structure allows researchers to systematically capture and analyze causal relationships in texts.

Other features: definition marks when a thesis defines something. sequencesGroup breaks down processes like recipes or procedures into steps. analogiesGroup captures comparisons. metaphorsGroup documents metaphorical language.

Below is an example showing how sequencesGroup breaks down processes like recipes or procedures into individual steps:

<sequencesGroup>
  <sequence>
    <phasesGroup>
      <newPhases>
        <phase>
          <textRef>
            <segment from="..." to="..."/>
          </textRef>
          <objectsGroup>
            <objectRef>...</objectRef>
          </objectsGroup>
          <phaseTemporality>
            <duration>...</duration>
            <repetitions>...</repetitions>
          </phaseTemporality>
          <paraphrasis>...</paraphrasis>
        </phase>
        <phase>...</phase>
      </newPhases>
    </phasesGroup>
  </sequence>
  <sequence>...</sequence>
</sequencesGroup>

The sequencesGroup is used to break down ordered series into their constituent phases. A sequence can be a chain of events (such as a historical narrative describing how events unfold), a reported or prescribed procedure (such as a recipe or set of instructions), or a mix of both.

Each sequence represents one complete ordered series (for example, a recipe for making lead white, or a historical account of a battle). Within each sequence, phasesGroup contains newPhases, which groups all the individual phases or steps.

Each phase represents one step or event in the sequence. For each phase, you can:

  • Point to the text that describes it using textRef and segment
  • List the objects or ingredients involved using objectsGroup and objectRef
  • Add timing information in phaseTemporality, such as how long it takes (duration) or how many times it's repeated (repetitions)
  • Provide a paraphrase to clarify what the phase means

This structure lets researchers capture detailed information about any ordered series found in historical sources, whether chains of events (like historical narratives), procedures (like recipes or medical treatments), or combinations of both. You can see both the order of phases and the specific details of each one.

This is a complete annotation of Theophrastus's recipe for making lead white (ψιμύθιον), as described in his work On Stones. The recipe is broken down into 7 phases, each capturing a specific step in the process. Below is the complete XML structure with detailed explanations.

Complete THESIS Structure

<THESIS xml:id="tlg0093_tlg004.T151115">
  <matchingPropositionsGroup>
    <matchingProposition propRef="#tlg0093_tlg004.P151115" comment="Theophrastus recipe" generalized="true"/>
  </matchingPropositionsGroup>
  <speakersGroup>
    <speaker name="./meta/ancient-lead-white_JOHD_people.xml#Theophrastus"/>
  </speakersGroup>
  <thesisType>
    <macroThemesGroup>
      <macroTheme macroThemeTag="./meta/ancient-lead-white_JOHD_indexing-tags.xml#historical"/>
      <macroTheme macroThemeTag="./meta/ancient-lead-white_JOHD_indexing-tags.xml#methodological"/>
    </macroThemesGroup>
    <microThemesGroup>
      <microTheme microThemeTag="./meta/ancient-lead-white_JOHD_indexing-tags.xml#chemistry"/>
      <microTheme microThemeTag="./meta/ancient-lead-white_JOHD_indexing-tags.xml#metallurgy"/>
      <microTheme microThemeTag="./meta/ancient-lead-white_JOHD_indexing-tags.xml#mineralogy"/>
    </microThemesGroup>
    <keywordsGroup>
      <keywordRef ref="#tlg0093_tlg004.K154432"/>
      <keywordRef ref="#tlg0093_tlg004.K154545" focus="2"/>
      <!-- ... more keywordRefs ... -->
    </keywordsGroup>
    <sequencesGroup>
      <sequence xml:id="tlg0093_tlg004.Q151115">
        <phasesGroup>
          <newPhases>
            <!-- 7 phases follow -->
          </newPhases>
        </phasesGroup>
      </sequence>
    </sequencesGroup>
  </thesisType>
  <text>
    <textRef>
      <segment from="./sources-segmented/tlg0093.tlg004.1st1K-grc1.xml#w1003410" to="./sources-segmented/tlg0093.tlg004.1st1K-grc1.xml#w1003458"/>
      <snippet>τίθεται γὰρ μόλυβδος ὑπὲρ ὄξους ἐν πίθοις ἡλίκον πλίνθος...</snippet>
    </textRef>
  </text>
  <paraphrasis>.Lead the size of a brick is placed over vinegar in (closed) jars, .They wait for lead to acquire thickness{, which happens in a maximum of 10 days}, .They open the jars, .They scrape away from the lead something {resembling a mold}, repeating this until they have consumed it completely, .They rub what has been scraped away in a grinding-bowl, .They always filter off what is being rubbed, .The final precipitate is lead white</paraphrasis>
</THESIS>

The Seven Phases Explained

Phase 1.1: Placing Lead Over Vinegar

<phase xml:id="tlg0093_tlg004.H151116">
  <textRef>
    <segment from="./sources-segmented/tlg0093.tlg004.1st1K-grc1.xml#w1003410"/>
    <segment from="./sources-segmented/tlg0093.tlg004.1st1K-grc1.xml#w1003412" to="./sources-segmented/tlg0093.tlg004.1st1K-grc1.xml#w1003418"/>
  </textRef>
  <paraphrasis>Lead the size of a brick is placed over vinegar in (closed) jars</paraphrasis>
  <objectsGroup>
    <object xml:id="tlg0093_tlg004.O155811">
      <keywordSubRef keywordsInSequence="1"/>
      <objectMeasuresGroup>
        <objectMeasure measurementType="./meta/ancient-lead-white_JOHD_indexing-tags.xml#size">
          <relativeTo><keywordSubRef keywordsInSequence="4"/></relativeTo>
        </objectMeasure>
      </objectMeasuresGroup>
      <objectFunction ingredient="true"/>
    </object>
    <object xml:id="tlg0093_tlg004.O164009">
      <keywordSubRef keywordsInSequence="2"/>
      <objectFunction ingredient="true"/>
    </object>
    <object xml:id="tlg0093_tlg004.O164046">
      <keywordSubRef keywordsInSequence="3"/>
      <objectQuantity unspecifiedPlurality="true"/>
      <objectFunction instrument="true"/>
    </object>
  </objectsGroup>
</phase>

This phase sets up the initial conditions. Three objects are identified: lead (ingredient, measured relative to a brick's size), vinegar (ingredient), and jars (instruments, plural). The objectFunction elements classify each object's role: ingredients vs. instruments.

Phase 1.2: Waiting for Lead to Acquire Thickness

<phase xml:id="tlg0093_tlg004.H151117">
  <textRef>...</textRef>
  <paraphrasis>They wait for lead to acquire thickness {, which happens in a maximum of 10 days}</paraphrasis>
  <phaseTemporality>
    <phaseDuration timeUnit="./meta/ancient-lead-white_JOHD_indexing-tags.xml#day" durationRange="-10">
      <relativeEnd>
        <fullyThemedTextRef>
          <textRef>
            <segment from="..." to="..."/>
            <snippet>μόλυβδος […] λάβῃ πάχος</snippet>
          </textRef>
          <macroThemesGroup>
            <macroTheme macroThemeTag="./meta/ancient-lead-white_JOHD_indexing-tags.xml#historical"/>
            <macroTheme macroThemeTag="./meta/ancient-lead-white_JOHD_indexing-tags.xml#physical"/>
          </macroThemesGroup>
        </fullyThemedTextRef>
      </relativeEnd>
      <proof>
        <elementRef ref="#tlg0093_tlg004.T180430"/>
      </proof>
    </phaseDuration>
  </phaseTemporality>
  <objectsGroup>
    <object xml:id="tlg0093_tlg004.O164447">
      <objectRef ref="#tlg0093_tlg004.O155811"/>
      <productOf phasesRef="1.1"/>
    </object>
  </objectsGroup>
</phase>

This phase includes temporal information: the duration is specified as "maximum of 10 days" using phaseDuration with durationRange="-10" (the minus sign indicates "up to"). The relativeEnd points to the text describing when the phase ends ("lead acquires thickness"). The productOf element links this phase's lead object back to the lead from phase 1.1, showing continuity.

Phase 1.3: Opening the Jars

A simple action phase: opening the jars. The objects reference the jars from phase 1.1 and the transformed lead from phase 1.2.

Phase 1.4: Scraping Away the Mold-Like Substance

<phase>
  <paraphrasis>They scrape away from the lead something {resembling a mold}, repeating this until they have consumed it completely</paraphrasis>
  <phaseTemporality>
    <phaseRepetition>
      <relativeEnd>
        <fullyThemedTextRef>
          <textRef>
            <segment from="..." to="..."/>
            <snippet>ὥσπερ εὐρῶτά τινα […] ἄν καταναλώσωσι</snippet>
          </textRef>
        </fullyThemedTextRef>
      </relativeEnd>
    </phaseRepetition>
  </phaseTemporality>
</phase>

This phase demonstrates repetition: the scraping action is repeated until the lead is completely consumed. The phaseRepetition element captures this iterative aspect, with relativeEnd pointing to the text that describes the stopping condition ("until they have consumed it completely").

Phase 1.5: Rubbing in a Grinding-Bowl

The scraped material is rubbed in a grinding-bowl. Objects reference the material from phase 1.4 and identify the grinding-bowl as an ingredient.

Phase 1.6: Filtering Off

This phase includes scholarly annotations: externalRef and multiple scholarContra elements document different interpretations of what "filtering off" means (decantation, levigation, etc.), showing how the schema captures scholarly debate.

Phase 1.7: Final Product

<phase>
  <paraphrasis>The final precipitate is lead white</paraphrasis>
  <objectsGroup>
    <object xml:id="tlg0093_tlg004.O221249">
      <keywordSubRef keywordsInSequence="7"/>
      <paraphrasis>The final precipitate of what resembles a mold after being scraped away from the lead, rubbed, and filtered off</paraphrasis>
      <productOf phasesRef="1.6">
        <fromMultipleObjects>
          <objectRef ref="#tlg0093_tlg004.O221025"/>
        </fromMultipleObjects>
        <productType portion="true"/>
      </productOf>
    </object>
  </objectsGroup>
  <containsPurpose value="true"/>
</phase>

The final phase identifies lead white as the end product. The productOf element shows this is a portion (productType portion="true") derived from the filtered material of phase 1.6. The containsPurpose value="true" marks this as the goal of the entire process.

Key Features Demonstrated

  • Object tracking: Objects are referenced across phases using objectRef, showing how materials transform and persist through the process.
  • Temporal information: Duration (phase 1.2) and repetition (phase 1.4) are captured with precise references to source text.
  • Product relationships: productOf elements link outputs to inputs, creating a chain from raw materials to final product.
  • Scholarly annotations: The schema can document different interpretations and scholarly debates (phase 1.6).
  • Cross-references: matchingPropositionSequence links this recipe to related recipes in other sources, enabling comparative analysis.
  • Complete traceability: Every phase references specific text segments, allowing researchers to trace each step back to the original source.

This annotation demonstrates how TheSu XML can capture complex multi-step processes with full detail, enabling systematic analysis of historical recipes, procedures, and technical instructions.

SUPPORT

Discourse components that provide context for theses, other supports, or other discourse elements. Supports can function argumentatively (as justifications), expositively (as clarifications), expansively (as elaborations), or contextually (as framing). They may be explicit or implicit, and can employ other elements while targeting the elements they relate to.

<SUPPORT xml:id="...">
  <targetsGroup>...</targetsGroup>
  <marker>...</marker>
  <speakersGroup>...</speakersGroup>
  <supportType>
    <supportFunction>
      <supportFunctionsGroup>
        <argumentation rank="..."/>
        <exposition rank="..."/>
        <expansion rank="..."/>
        <contextualization rank="..."/>
      </supportFunctionsGroup>
    </supportFunction>
    <supportForm formTag="..."/>
  </supportType>
  <employedElements>...</employedElements>
  <text>...</text>
  <paraphrasis>...</paraphrasis>
</SUPPORT>

Child Elements

  • targetsGroup — Links the support to the THESIS, MISC, or SUPPORT elements it relates to. May contain omittedTargets if some targets are not annotated.
  • marker — Points to words or phrases in the text (e.g., "because", "namely", "therefore", "however") that signal how the support functions. Used for explicit supports where these markers are present.
  • speakersGroup — Identifies who is speaking or presenting the support. Contains one or more speaker elements, each with optional @rank attribute to show how certain the identification is when it's unclear.
  • supportType — Describes what the support does and how it's structured. Contains supportFunction with supportFunctionsGroup (listing all four possible functions: argumentation, exposition, expansion, contextualization, each with optional @rank to indicate prominence), and supportForm elements (describing the structural form, e.g., "particular-illustration", "analogy", "deductive reasoning").
  • employedElements — References THESIS, MISC, or SUPPORT elements that appear in the support's text and are used as part of its explanation (e.g., theses used as premises in an argument). May contain omittedEmployedElements if some employed elements are not annotated.
  • text — Points to one or more segments of the original source text via textRef elements, each containing segment elements that identify the text spans. May include differences for variant readings and uncertainties for corrupt or uncertain text.
  • paraphrasis — Provides a clear, concise paraphrase of what the support means in the corpus language. Required for implicit or extrinsic supports, where the actual text differs from what the support represents. Must match all other annotated details.

Supports can have multiple functions simultaneously, and each function can be ranked to show its prominence. The supportForm element describes the structural form the support takes.

This example shows a complete SUPPORT element with multiple functions. The support uses the concrete example of hair turning white with age to justify a general principle about how things become like what is present to them.

<SUPPORT xml:id="tlg0059_tlg020.S135633">
  <targetsGroup>
    <target ref="#tlg0059_tlg020.T204411"/>
  </targetsGroup>
  <marker>
    <textRef>
      <segment from="./sources-segmented/tlg0059.tlg020.perseus-grc2.xml#w1004969"/>
      <segment from="./sources-segmented/tlg0059.tlg020.perseus-grc2.xml#w1004971"/>
      <snippet>τοίνυν […] νῦν</snippet>
    </textRef>
  </marker>
  <speakersGroup>
    <speaker name="./meta/ancient-lead-white_JOHD_people.xml#Socrates"/>
  </speakersGroup>
  <supportType>
    <supportFunction>
      <supportFunctionsGroup>
        <argumentation rank="1"/>
        <exposition rank="2"/>
        <expansion/>
        <contextualization/>
      </supportFunctionsGroup>
    </supportFunction>
    <supportForm formTag="./meta/ancient-lead-white_JOHD_relationships-tags.xml#particular-illustration"/>
  </supportType>
  <employedElements>
    <elementRef ref="#tlg0059_tlg020.T200422"/>
  </employedElements>
  <text>
    <textRef>
      <segment from="./sources-segmented/tlg0059.tlg020.perseus-grc2.xml#w1004945" to="./sources-segmented/tlg0059.tlg020.perseus-grc2.xml#w1004964"/>
      <segment from="./sources-segmented/tlg0059.tlg020.perseus-grc2.xml#w1004968" to="./sources-segmented/tlg0059.tlg020.perseus-grc2.xml#w1004972"/>
      <snippet>ἀλλʼ ὅταν δή , ὦ φίλε , τὸ γῆρας αὐταῖς ταὐτὸν τοῦτο χρῶμα ἐπαγάγῃ , τότε ἐγένοντο οἷόνπερ τὸ παρόν , λευκοῦ παρουσίᾳ λευκαί . τοῦτο τοίνυν ἐρωτῶ νῦν δή ,</snippet>
    </textRef>
  </text>
  <paraphrasis>"When old age has brought (to hair) the (white) colour, (the hair) becomes like that which is present, i.e., white by presence of white" JUSTIFIES/explains) "( When something is present, if this is present in a certain way, that which possesses (what is present to it) is such as what is present )"</paraphrasis>
</SUPPORT>

Support functions and form:

The supportFunctionsGroup shows this support has two functions: argumentation with rank="1" (primary function) and exposition with rank="2" (secondary function). This means the support primarily functions as an argumentative justification, but also serves to clarify or explain the target thesis. The functions expansion and contextualization are present but have no rank specified, meaning they default to rank="4" (no particular relevance) and are not active functions for this support.

The supportForm element specifies formTag="particular-illustration", indicating that this support takes the structural form of a particular illustration—it uses a specific concrete example (hair turning white with age) to support a general principle.

Support relationships:

  • Employed element: The support uses the thesis "When old age has brought (to hair) the (white) colour, (the hair) becomes like that which is present, i.e., white by presence of white" as the concrete example that serves as its premise.
  • Target: The support targets the thesis "When something is present, if this is present in a certain way, that which possesses (what is present to it) is such as what is present", providing argumentative justification (rank 1) and clarification (rank 2) for this general principle through the particular illustration of hair turning white.

The supportType element describes what a support does and how it's structured. It contains two main parts: supportFunction (what the support does) and supportForm (how it's structured).

Support Functions: The supportFunctionsGroup always contains all four possible functions: argumentation, exposition, expansion, and contextualization. A support can have one or more of these functions active at the same time. Each function can have a @rank attribute to show how prominent it is:

  • Rank "1" — Primary or most prominent function
  • Rank "2" — Secondary function
  • Rank "3" — Tertiary function
  • No rank specified — Defaults to rank "4", meaning the function is not particularly relevant to this support

Assigning ranks to support functions is an interpretative decision made by the annotator based on analysis of the text. The annotator determines which function is most prominent (rank="1"), which is secondary (rank="2"), and so on, by examining how the support is structured, what it emphasises, and how it relates to its targets. This process requires close reading to identify both explicit signals and implicit patterns, following established scholarly practices in history and philology. The schema frames this as the annotator's task, acknowledging that annotation involves interpretation rather than objective fact-finding.

For example, a support might be primarily argumentative (rank="1") but also serve an expository function (rank="2"), meaning it both justifies a claim and clarifies it. The other two functions (expansion and contextualization) would be present but have no rank, indicating they're not active for this support.

Support Form: The supportForm element describes the structural form the support takes—the way it presents its argument or explanation. Common forms include:

  • "particular-illustration" — Using a specific concrete example to support a general principle
  • "analogy" — Drawing comparisons between different things
  • "general-to-particular" — Deriving a specific conclusion from a general principle
  • "authority" — Appealing to authoritative sources or expert opinions

A support can have multiple supportForm elements if it uses more than one structural approach.

<supportType>
  <supportFunction>
    <supportFunctionsGroup>
      <argumentation rank="1"/>
      <exposition rank="2"/>
      <expansion/>
      <contextualization/>
    </supportFunctionsGroup>
  </supportFunction>
  <supportForm formTag="./meta/ancient-lead-white_JOHD_relationships-tags.xml#particular-illustration"/>
</supportType>

In this example, the support functions primarily as argumentation (rank="1") and secondarily as exposition (rank="2"), using the structural form of a particular illustration to make its point.

This example shows a complete SUPPORT element with cumulative functions—it has both exposition (rank="1") and argumentation (rank="2") working together. The support uses a detailed recipe to both explain what "technical means" means and justify that lead white production is indeed a technical process.

<SUPPORT xml:id="tlg0093_tlg004.S153946">
  <targetsGroup>
    <target ref="#tlg0093_tlg004.T170246"/>
  </targetsGroup>
  <marker>
    <textRef>
      <segment from="./sources-segmented/tlg0093.tlg004.1st1K-grc1.xml#w1003411"/>
      <snippet>γὰρ</snippet>
    </textRef>
  </marker>
  <speakersGroup>
    <speaker name="./meta/ancient-lead-white_JOHD_people.xml#Theophrastus"/>
  </speakersGroup>
  <supportType>
    <supportFunction>
      <supportFunctionsGroup>
        <argumentation rank="2"/>
        <exposition rank="1"/>
        <expansion/>
        <contextualization/>
      </supportFunctionsGroup>
    </supportFunction>
    <supportForm formTag="./meta/ancient-lead-white_JOHD_relationships-tags.xml#particular-illustration"/>
  </supportType>
  <employedElements>
    <elementRef ref="#tlg0093_tlg004.T151115"/>
  </employedElements>
  <text>
    <textRef>
      <segment from="./sources-segmented/tlg0093.tlg004.1st1K-grc1.xml#w1003410" to="./sources-segmented/tlg0093.tlg004.1st1K-grc1.xml#w1003458"/>
      <snippet>τίθεται γὰρ μόλυβδος ὑπὲρ ὄξους ἐν πίθοις ἡλίκον πλίνθος . ὅταν δὲ λάβῃ πάχος , λαμβάνει δὲ μάλιστα ἐν ἡμέραις δέκα , τότʼ ἀνοίγουσιν , εἶτʼ ἀποξύουσιν ὥσπερ εὐρῶτά τινα ἀπʼ αὐτοῦ , καὶ πάλιν , ἕως ἄν καταναλώσωσι . τὸ δʼ ἀποξυόμενον ἐν τριπτῆρι τρίβουσι καὶ ἀφηθοῦσιν ἀεὶ , τὸ δʼ ἔσχατον ὑφιστάμενόν ἐστι τὸ ψιμύθιον .</snippet>
    </textRef>
  </text>
  <paraphrasis>"Lead the size of a brick is placed over vinegar in (closed) jars, .They wait for lead to acquire thickness{, which happens in a maximum of 10 days}, .They open the jars, .They scrape away from the lead something {resembling a mold}, repeating this until they have consumed it completely, .They rub what has been scraped away in a grinding-bowl, .They always filter off what is being rubbed, .The final precipitate is lead white" EXPLAINS/justifies) " Lead white comes about by technical means"</paraphrasis>
</SUPPORT>

Cumulative functions:

Determining that exposition has rank="1" and argumentation has rank="2" is an interpretative decision based on textual analysis. In this case, the annotator judged that the support primarily functions to explain/specify (exposition) because it provides detailed recipe steps that show HOW lead white comes about by technical means. The argumentative function (justifying that it IS a technical process) is present but secondary, as the recipe serves primarily to clarify the process rather than to prove it. This interpretation is based on the text's structure—the detailed step-by-step presentation emphasises explanation over justification—but different annotators might reasonably emphasise different aspects. This interpretative process, which captures both explicit content and implicit reasoning through close reading, follows established practices in history and philology, making scholarly interpretations visible and traceable through the annotations.

  • The exposition function has rank="1" (primary), meaning the support's main purpose is to explain in what sense "lead white comes about by technical means"—it specifies how this happens by providing concrete recipe steps that show the technical process involved.
  • The argumentation function has rank="2" (secondary), meaning the support also justifies the claim that lead white production is a technical process. The detailed recipe serves as evidence that this is indeed a technical procedure, not a natural or spontaneous process.
  • The other two functions (expansion and contextualization) have no rank specified, defaulting to rank="4", meaning they are not active functions for this support.

Support form:

The supportForm element specifies formTag="particular-illustration", indicating that this support uses the structural form of a particular illustration—it uses a specific concrete example (the detailed recipe) to support a general principle (that lead white comes about by technical means).

Support relationships:

  • Employed element: The recipe sequence (#tlg0093_tlg004.T151115) is used as the concrete example that serves as both explanation and justification.
  • Target: The general claim "Lead white comes about by technical means" (#tlg0093_tlg004.T170246) is what the support explains and justifies through the recipe.

PROPOSITION

Abstract ideas that link multiple theses expressing similar core concepts, allowing comparison across texts. Propositions are not tied to specific text spans and include paraphrases describing the abstract idea. Theses link to propositions that match them, whether within the same source or across multiple sources.

When a THESIS element is linked to a PROPOSITION, the thesis expresses an instance of that abstract idea. Multiple theses can link to the same proposition. The matching doesn't have to be exact—propositions can be similar enough to be connected, reflecting the annotator's interpretation of how ideas relate.

<PROPOSITION xml:id="...">
  <relationships>...</relationships>
  <propositionType>...</propositionType>
  <paraphrasis>...</paraphrasis>
</PROPOSITION>

Child Elements

  • relationships — Links the proposition to other PROPOSITION elements via semantic relationships. Contains entailment for logical entailment relationships, similarity for similar meanings, and opposition for opposed meanings.
  • propositionType — Contains thematic classification and formal features, similar to thesisType. Includes macroThemesGroup for broad themes (e.g., "historical", "axiological"), microThemesGroup for detailed themes (e.g., "military", "psychological"), keywordsGroup for keywords (which can be annotated directly here, unlike in THESIS where they must reference separately annotated keywords), definition if the proposition has definitional structure, sequencesGroup for processual analyses (recipes, procedures, chains of events), etiologiesGroup for causal structures, analogiesGroup for analogical relationships, and metaphorsGroup for metaphorical expressions.
  • paraphrasis — Provides a clear, concise paraphrase of the proposition's semantic content in the corpus language. Since propositions are not linked to specific text spans, the paraphrasis element is required to describe the abstract idea. Must be consistent with all other annotated details (polarity, thematic classification, formal structure, etc.).

Theses link to propositions through the matchingPropositionsGroup element, which is a child element of THESIS. This element contains one or more matchingProposition references that point to PROPOSITION elements expressing similar ideas.

<THESIS xml:id="...">
  <matchingPropositionsGroup>
    <matchingProposition propRef="#proposition_id"/>
    <matchingProposition propRef="#another_proposition_id"/>
  </matchingPropositionsGroup>
  ...</THESIS>

This PROPOSITION connects two theses from different sources. The proposition "Lead white is a cooling substance" links theses expressing the same idea about lead white's cooling properties.

<PROPOSITION xml:id="w0007112.P664643">
  <propositionType>
    <macroThemesGroup>
      <macroTheme macroThemeTag="./meta/ancient-lead-white_JOHD_indexing-tags.xml#physical"/>
    </macroThemesGroup>
    <microThemesGroup>
      <microTheme microThemeTag="./meta/ancient-lead-white_JOHD_indexing-tags.xml#mineralogy"/>
      <microTheme microThemeTag="./meta/ancient-lead-white_JOHD_indexing-tags.xml#chemistry"/>
    </microThemesGroup>
    <keywordsGroup>
      <keywordRef wordsInParaphrasis="1-2" namely="./meta/ancient-lead-white_JOHD_indexing-tags.xml#lead-white"/>
      <keywordRef wordsInParaphrasis="5" namely="./meta/ancient-lead-white_JOHD_indexing-tags.xml#cooling_a"/>
    </keywordsGroup>
    <definition/>
    <sequencesGroup/>
    <etiologiesGroup/>
    <analogiesGroup/>
  </propositionType>
  <paraphrasis>Lead white is a cooling substance</paraphrasis>
</PROPOSITION>

The keywordsGroup element in PROPOSITION works differently from THESIS. Since propositions don't link to specific text spans, keywords must be tagged directly using @namely and linked to words in the paraphrasis.

@namely attribute: Required for PROPOSITION keywords. Tags the keyword directly and links to an authority record (e.g., ./meta/ancient-lead-white_JOHD_indexing-tags.xml#lead-white). This ensures consistent keyword tagging across the corpus.

@wordsInParaphrasis attribute: Required for PROPOSITION keywords. Specifies which words in the paraphrasis correspond to the keyword. Words are numbered in order of appearance, separated by spaces. For example, in the paraphrasis "Lead white is a cooling substance":

  • Word 1: "Lead"
  • Word 2: "white"
  • Word 3: "is"
  • Word 4: "a"
  • Word 5: "cooling"
  • Word 6: "substance"

In the example above, wordsInParaphrasis="1-2" refers to "Lead white", and wordsInParaphrasis="5" refers to "cooling". This precise mapping links keywords to their positions in the paraphrasis, since propositions have no source text spans to reference.

Two theses linking to this proposition:

  • THESIS tlg0656_tlg001.T161627 from De materia medica (Περὶ ὕλης ἰατρικῆς) by Dioscorides: "(Lead white)'s property is: cooling, adhesive-malleability-inducing, softening..." — links to this proposition with @partial="true", indicating the thesis matches part of the proposition (the thesis lists multiple properties, while the proposition focuses on cooling).
    <THESIS xml:id="tlg0656_tlg001.T161627">
      <matchingPropositionsGroup>
        <matchingProposition propRef="#w0007112.P664643" partial="true"/>
      </matchingPropositionsGroup>
      ...
      <speakersGroup>
        <speaker name="./meta/ancient-lead-white_JOHD_people.xml#Dioscorides-Pedanius"/>
      </speakersGroup>
      ...
      <text>
        <snippet>δύναμιν δὲ ἔχει ψυκτικήν , ἐμπλαστικήν , μαλακτικήν , πληρωτικήν , λεπτυντικήν , ἔτι δὲ πράως κατασταλτικὴν ὑπεροχῶν καὶ κατουλωτικήν , μειγνύμενον κηρωταῖς καὶ λιπαραῖς ἐμπλάστροις καὶ τροχίσκοις</snippet>
      </text>
      <paraphrasis>(Lead white)'s property is: cooling, adhesive-malleability-inducing, softening, filling, thinning, moderately controlling of excrescences, and – when mixed with cerates, fatty plasters and lozenges – cicatrising</paraphrasis>
    </THESIS>
  • THESIS tlg0007_tlg112.T145739 from Quaestiones convivales (Συμποσιακῶν προβλημάτων βιβλία) by Plutarch: "Lead white is the most cooling of deadly drugs" — links to this proposition with @generalized="true" and @partial="true", indicating the proposition is more general (it doesn't specify "most cooling" or "deadly drugs") and corresponds to only part of the thesis's meaning, with the thesis including additional information.
    <THESIS xml:id="tlg0007_tlg112.T145739">
      <matchingPropositionsGroup>
        <matchingProposition propRef="#w0007112.P664643" generalized="true" partial="true"/>
      </matchingPropositionsGroup>
      ...
      <speakersGroup>
        <speaker name="./meta/ancient-lead-white_JOHD_people.xml#Plutarchus"/>
      </speakersGroup>
      ...
      <text>
        <snippet>τὸ ψυκτικώτατον τῶν θανασίμων φαρμάκων […] ψιμύθιον</snippet>
      </text>
      <paraphrasis>Lead white is the most cooling of deadly drugs</paraphrasis>
    </THESIS>

Both theses reference the same PROPOSITION through their matchingPropositionsGroup elements.

The matchingProposition element uses attributes to specify how a thesis relates to the proposition it references. The matching doesn't have to be exact—propositions can be similar enough to be usefully connected, and attributes indicate the nature of the relationship. Multiple attributes can be used together when a relationship has multiple characteristics.

  • @partial — The proposition corresponds to only part of the thesis's meaning. The thesis includes additional information beyond what the proposition covers. Example: thesis "temperance and resolution benefit our intellect and well-being..." linked to proposition "temperance benefits our intellect".
  • @extended — The proposition includes the thesis while also providing additional information. The proposition is broader or more comprehensive than the thesis. Example: thesis "temperance benefits our intellect" linked to proposition "temperance and resolution benefit our intellect and well-being...".
  • @quoted — The proposition frames the thesis in a quotation or report. The proposition presents the thesis as quoted or reported speech. Example: thesis "temperance benefits our intellect" linked to proposition "Aristotle said that temperance benefits our intellect".
  • @extracted — The thesis is a quotation or report, and the proposition matches the quoted sentence rather than the quotation itself. This extracts the quoted thesis from the quotation that frames it. Example: thesis "Aristotle said that temperance benefits our intellect" linked to proposition "temperance benefits our intellect".
  • @specified — The proposition specifies the thesis in detail, adding information that makes it potentially encompassed by the thesis. Example: thesis "temperance benefits our intellect" linked to proposition "the virtue of temperance usually benefits our intellect by removing superfluous distractions".
  • @generalized — The proposition generalises the thesis, expressing it in broader terms that could encompass it. Example: thesis "temperance benefits our intellect" linked to proposition "certain virtues can benefit our mind".
  • @altered — The proposition is similar to the thesis but diverges in some details. Meanings are compatible but differ in specific aspects. Example: thesis "temperance benefits our intellect" linked to proposition "discipline and orderliness benefit our intellect".

Attributes can be used cumulatively. For example, a thesis might be linked to a proposition that is both extended, extracted, and altered relative to it, requiring multiple attributes to be set to "true".

PROPOSITION w0007112.P664643: "Lead white is a cooling substance"

Two theses linking to this proposition with different attribute combinations:

<THESIS xml:id="tlg0656_tlg001.T161627">
  <matchingPropositionsGroup>
    <matchingProposition propRef="#w0007112.P664643" partial="true"/>
  </matchingPropositionsGroup>
  ...
  <paraphrasis>(Lead white)'s property is: cooling, adhesive-malleability-inducing, softening, filling, thinning, moderately controlling of excrescences, and – when mixed with cerates, fatty plasters and lozenges – cicatrising</paraphrasis>
</THESIS>

Dioscorides, De materia medica: Uses @partial="true" because the proposition corresponds to only part of the thesis's meaning (the thesis lists multiple properties, while the proposition focuses on cooling).

<THESIS xml:id="tlg0007_tlg112.T145739">
  <matchingPropositionsGroup>
    <matchingProposition propRef="#w0007112.P664643" generalized="true" partial="true"/>
  </matchingPropositionsGroup>
  ...
  <paraphrasis>Lead white is the most cooling of deadly drugs</paraphrasis>
</THESIS>

Plutarch, Quaestiones convivales: Uses @generalized="true" and @partial="true" because the proposition is more general (it doesn't specify "most cooling" or "deadly drugs") and corresponds to only part of the thesis's meaning, with the thesis including additional information.

When propositions contain sequences (processes, recipes, or chains of events), theses can map their own sequence phases to the proposition's phases. This shows how different sources describe the same sequence. Some sources include more detail, others less. Some present phases in different order, or describe them differently. For procedural sequences like recipes or instructions, this phase mapping shows how different sources present the same procedure step by step—which steps they include, how they order them, and what methods they use. By mapping phases, you can see how sources correspond, differ, or transform the same sequence, and how knowledge changed as it moved between sources.

When a THESIS with a sequence links to a PROPOSITION that also contains a sequence, the phases can be mapped analytically. This shows how different sources describe the same process step-by-step.

At the sequence level, the THESIS sequence references the PROPOSITION sequence using matchingPropositionSequence, which points to the proposition's sequence via @sequenceRef. At the phase level, individual phases map to proposition phases using matchingPropositionPhases, which references specific phases via @phasesRef (e.g., "1.1" for the first phase, "1.2-1.4" for a range).

<THESIS xml:id="...">
  <matchingPropositionsGroup>
    <matchingProposition propRef="#proposition_id" .../>
  </matchingPropositionsGroup>
  ...
  <thesisType>
    <sequencesGroup>
      <sequence xml:id="thesis_sequence_id">
        <matchingPropositionSequence sequenceRef="#proposition_sequence_id" comment="..."/>
        <phasesGroup>
          <newPhases>
            <phase>
              <matchingPropositionPhases phasesRef="1.1" generalized="true"/>
              ...
            </phase>
          </newPhases>
        </phasesGroup>
      </sequence>
    </sequencesGroup>
  </thesisType>
</THESIS>

The matchingPropositionPhases element can use the same attributes as matchingProposition (generalized, extended, specified, altered, partial) to describe how individual phases match.

Phase references in @phasesRef use a coded string format that identifies phases by their position in the proposition sequence. The format is "X.Y" where X refers to the phasesGroup (starting from 1) and Y refers to the phase's position within that phasesGroup's newPhases (also starting from 1). For example, "1.1" refers to the first phase in the first phasesGroup, "1.2" to the second phase in the first phasesGroup, and "1.2-1.4" refers to phases 1.2, 1.3, and 1.4. The proposition sequence referenced by matchingPropositionSequence determines which sequence these references point to.

<PROPOSITION xml:id="tlg0093_tlg004.P151115">
  <propositionType>
    <sequencesGroup>
      <sequence xml:id="tlg0093_tlg004.Q151116">
        <phasesGroup>
          <newPhases>
            <phase>...</phase> <!-- Phase 1.1 -->
            <phase>...</phase> <!-- Phase 1.2 -->
            <phase>...</phase> <!-- Phase 1.3 -->
            <phase>...</phase> <!-- Phase 1.4 -->
            <phase>...</phase> <!-- Phase 1.5 -->
            <phase>...</phase> <!-- Phase 1.6 -->
            <phase>...</phase> <!-- Phase 1.7 -->
          </newPhases>
        </phasesGroup>
      </sequence>
    </sequencesGroup>
  </propositionType>
</PROPOSITION>

When a thesis phase uses matchingPropositionPhases phasesRef="1.1", it maps to the first phase in this proposition sequence. Using phasesRef="1.4" maps to the fourth phase, and phasesRef="1.2-1.4" maps to phases 1.2, 1.3, and 1.4.

This structure shows how different authors describe the same procedures with varying levels of detail, different ordering, or alternative approaches.

PROPOSITION tlg0093_tlg004.P151115:
Theophrastus's recipe for producing lead white, broken down into 7 phases.

<PROPOSITION xml:id="tlg0093_tlg004.P151115">
  <propositionType>
    ...
    <sequencesGroup>
      <sequence xml:id="tlg0093_tlg004.Q151116">
        <phasesGroup>
          <newPhases>
            <phase>
              <microThemedFreeText>
                <freeText>Place a piece of lead over vinegar in closed jars</freeText>
              </microThemedFreeText>
              <objectsGroup>
                <object xml:id="tlg0093_tlg004.O151115">
                  <keywordSubRef keywordsInSequence="1"/>
                  <objectFunction ingredient="true"/>
                </object>
                ...
              </objectsGroup>
            </phase>
            <phase>
              <microThemedFreeText>
                <freeText>Wait for lead to acquire thickness</freeText>
              </microThemedFreeText>
              <objectsGroup>
                <object xml:id="tlg0093_tlg004.O151118">
                  <objectRef ref="#tlg0093_tlg004.O151115"/>
                  <productOf phasesRef="1.1"/>
                </object>
                ...
              </objectsGroup>
            </phase>
            <phase>
              <microThemedFreeText>
                <freeText>Open the jars</freeText>
              </microThemedFreeText>
              <objectsGroup>
                <object xml:id="tlg0093_tlg004.O151111" implicit="true">
                  <objectRef ref="#tlg0093_tlg004.O151118"/>
                  <paraphrasis>lead after acquiring thickness</paraphrasis>
                  <productOf phasesRef="1.2">
                    <productType transformation="true"/>
                  </productOf>
                </object>
              </objectsGroup>
            </phase>
            <phase>
              <microThemedFreeText>
                <freeText>Scrape away from the lead the substance that has formed on its surface, repeating this until you have consumed it completely</freeText>
              </microThemedFreeText>
              <phaseTemporality>
                <phaseRepetition>
                  <relativeEnd>
                    <fullyThemedFreeText>
                      <freeText>until you have consumed it completely</freeText>
                    </fullyThemedFreeText>
                  </relativeEnd>
              </phaseRepetition>
            </phaseTemporality>
              <objectsGroup>
                <object xml:id="tlg0093_tlg004.O151112">
                  <objectRef ref="#tlg0093_tlg004.O151111"/>
                </object>
                ...
              </objectsGroup>
            </phase>
            <phase>
              <microThemedFreeText>
                <freeText>Grind what you have scraped away</freeText>
              </microThemedFreeText>
              <objectsGroup>
                <object xml:id="tlg0093_tlg004.O151114">
                  <objectRef ref="#tlg0093_tlg004.O151113"/>
                  <paraphrasis>the substance that formed on lead's surface after being scraped away from it</paraphrasis>
                  <productOf phasesRef="1.4"/>
                </object>
              </objectsGroup>
            </phase>
            <phase>
              <microThemedFreeText>
                <freeText>Filter off what you are grinding</freeText>
              </microThemedFreeText>
              <objectsGroup>
                <object xml:id="tlg0093_tlg004.O151120">
                  <objectRef ref="#tlg0093_tlg004.O151114"/>
                  <paraphrasis>the substance that formed on lead's surface after being scraped away from it and ground</paraphrasis>
                  <productOf phasesRef="1.5"/>
                </object>
              </objectsGroup>
            </phase>
            <phase>
              <microThemedFreeText>
                <freeText>The final precipitate is lead white</freeText>
              </microThemedFreeText>
              <objectsGroup>
                <object xml:id="tlg0093_tlg004.O151121">
                  <keywordSubRef keywordsInSequence="4"/>
                  <paraphrasis>the subastance that formed on lead's surface after being scraped away from it, ground, and filtered off</paraphrasis>
                  <productOf phasesRef="1.6">
                    <fromMultipleObjects>
                      <objectRef ref="#tlg0093_tlg004.O221025"/>
                    </fromMultipleObjects>
                    <productType portion="true"/>
                  </productOf>
                </object>
              </objectsGroup>
              <containsPurpose value="true"/>
            </phase>
          </newPhases>
        </phasesGroup>
        ...
      </sequence>
      ...
    </sequencesGroup>
    ...
  </propositionType>
  ...
</PROPOSITION>

This proposition contains a single sequence with 7 phases. Each phase describes one step in the recipe. The first phase sets up the process: placing lead over vinegar in closed jars. The second phase waits for the lead to transform. The third opens the jars. The fourth scrapes away the substance that formed, repeating until the lead is consumed. The fifth grinds what was scraped. The sixth filters it. The seventh identifies the final product: lead white.

Each phase includes a description of what happens (in microThemedFreeText) and lists the objects involved (in objectsGroup). Objects can be ingredients, instruments, or products from earlier phases. Some phases show how objects transform or connect across phases using productOf elements that reference earlier phases. The final phase marks the purpose of the entire sequence using containsPurpose value="true".

This proposition was modelled after a THESIS in Theophrastus's On Stones (Περὶ λίθων), which describes the recipe for producing lead white. Theophrastus's thesis itself references this proposition.

Theses linking to this proposition:

Let us focus on two theses from different authors that link to this proposition, each mapping their sequence phases to the proposition's phases. The mapping mechanism uses matchingPropositionSequence elements at the sequence level and matchingPropositionPhases elements at each phase level. Attributes describe how the thesis relates to the proposition: @generalized means the thesis is more general, @specified means it adds detail, @extended means it adds steps, @altered means it changes something, and @partial means it covers only part of the proposition. When a phase has no match, phasesRef="/" indicates this.

1. Dioscorides's recipe from De materia medica

Dioscorides's thesis links to this proposition with @extended="true" @specified="true" @generalized="true" @altered="true", indicating it expands, adds detail, generalises, and modifies the Theophrastus procedure. The thesis has 13 phases compared to the proposition's 7. Dioscorides, indeed, adds steps not present in Theophrastus.

<THESIS>
  <matchingPropositionsGroup>
    <matchingProposition propRef="#tlg0093_tlg004.P151115" extended="true" specified="true" generalized="true" altered="true"/>
  </matchingPropositionsGroup>
  <thesisType>
    <sequencesGroup>
      <sequence>
        <matchingPropositionSequence sequenceRef="#tlg0093_tlg004.Q151116"/>
        <phasesGroup>
          <newPhases>
            <phase>
              <matchingPropositionPhases phasesRef="1.1" extended="true" generalized="true"/>
              <paraphrasis>(In summer, ) pour most pungent vinegar into a broad-mouthed jar</paraphrasis>
            </phase>
            <phase>
              <matchingPropositionPhases phasesRef="/"/>
              <paraphrasis>Place a leaden brick on a lattice of reeds</paraphrasis>
            </phase>
            <phase>
              <matchingPropositionPhases phasesRef="1.1" extended="true" generalized="true"/>
              <paraphrasis>Lay the lattice of reeds with the leaden brick onto the mouth of the pot (i.e., of the broad-mouthed jar)</paraphrasis>
            </phase>
            <phase>
              <matchingPropositionPhases phasesRef="1.1" extended="true" generalized="true"/>
              <paraphrasis>Above, throw over covers (to prevent vinegar from evaporating)</paraphrasis>
            </phase>
            <phase>
              <matchingPropositionPhases phasesRef="1.2" altered="true"/>
              <paraphrasis>Wait until (the leaden brick), crumbling down, has strained through (the lattice)</paraphrasis>
            </phase>
            <phase>
              <matchingPropositionPhases phasesRef="/"/>
              <paraphrasis>Decant away the clean fluid that floats on top (of the vinegar after the leaden brick has crumbled down in it)</paraphrasis>
            </phase>
            <phase>
              <matchingPropositionPhases phasesRef="/"/>
              <paraphrasis>Pour the (remaining) viscous fluid into a vessel</paraphrasis>
            </phase>
            ...
          </newPhases>
        </phasesGroup>
        ...
      </sequence>
      ...
    </sequencesGroup>
    ...
  </thesisType>
  ...
</THESIS>

Phase 1 maps to proposition phase 1.1 with @extended="true" @generalized="true": the thesis adds detail about timing and vinegar type not present in the Theophrastus proposition, as shown in the paraphrasis "(In summer, ) pour most pungent vinegar into a broad-mouthed jar". Phase 2 ("Place a leaden brick on a lattice of reeds") has no match in the proposition (phasesRef="/"). Phase 3 ("Lay the lattice of reeds with the leaden brick onto the mouth of the pot") maps to proposition phase 1.1 with @extended="true" @generalized="true". Phase 4 ("Above, throw over covers") also maps to proposition phase 1.1 with @extended="true" @generalized="true". Phase 5 ("Wait until (the leaden brick), crumbling down, has strained through (the lattice)") maps to proposition phase 1.2 with @altered="true". Phases 6 ("Decant away the clean fluid that floats on top") and 7 ("Pour the (remaining) viscous fluid into a vessel") have no match in the proposition (phasesRef="/"). Many of the 13 phases remain unlinked to the proposition, including steps like decanting fluid and drying under the sun.

2. Plutarch's simplified description from Quaestiones convivales

Plutarch's thesis links to this proposition with @partial="true" @extended="true" @specified="true" @generalized="true" @altered="true", indicating it covers only part of the proposition, expands it, adds detail, generalises, and modifies it. The thesis has only 2 phases compared to the proposition's 7, condensing multiple proposition phases into single thesis phases.

<THESIS>
  <matchingPropositionsGroup>
    <matchingProposition propRef="#tlg0093_tlg004.P151115" partial="true" extended="true" specified="true" generalized="true" altered="true"/>
  </matchingPropositionsGroup>
  <thesisType>
    <sequencesGroup>
      <sequence>
        <matchingPropositionSequence sequenceRef="#tlg0093_tlg004.Q151116"/>
        <phasesGroup>
          <newPhases>
            <phase>
              <matchingPropositionPhases phasesRef="1.1,1.4" extended="true" specified="true" altered="true" partial="true"/>
              <paraphrasis>Lead is rubbed with vinegar</paraphrasis>
            </phase>
            <phase>
              <matchingPropositionPhases phasesRef="1.4" extended="true" generalized="true" altered="true"/>
              <paraphrasis>It sends forth the most cooling of deadly drugs, i.e., lead white</paraphrasis>
            </phase>
          </newPhases>
        </phasesGroup>
        ...
      </sequence>
      ...
    </sequencesGroup>
    ...
  </thesisType>
  ...
</THESIS>

The first thesis phase ("Lead is rubbed with vinegar") maps to proposition phases 1.1 and 1.4. Proposition phase 1.1 is "Place a piece of lead over vinegar in closed jars", and proposition phase 1.4 is "Scrape away from the lead the substance that has formed on its surface, repeating this until you have consumed it completely". Plutarch's phase combines these two proposition phases into a single description. The attributes explain how the proposition phases relate to Plutarch's phase: @extended="true" means the proposition phases include Plutarch's phase while also providing additional information (the proposition phases describe placing lead and scraping separately, while Plutarch combines them into "rubbed"); @specified="true" means the proposition phases specify Plutarch's phase in detail (the proposition phases add information about jars, closed containers, and the scraping process); @altered="true" means the proposition phases are similar to Plutarch's phase but diverge in details (the proposition phases describe separate steps while Plutarch describes a single action); and @partial="true" means the proposition phases correspond to only part of Plutarch's phase's meaning (Plutarch's phase covers the combined action, while the proposition phases cover only the separate steps).

The second thesis phase ("It sends forth the most cooling of deadly drugs, i.e., lead white") maps to proposition phase 1.4 ("Scrape away from the lead the substance that has formed on its surface, repeating this until you have consumed it completely"). Plutarch's phase describes the outcome rather than the process. The attributes explain how the proposition phase relates to Plutarch's phase: @extended="true" means the proposition phase includes Plutarch's phase while also providing additional information (the proposition phase describes the scraping process, while Plutarch describes the result); @generalized="true" means the proposition phase generalises Plutarch's phase (the proposition phase describes the process in broader terms, while Plutarch specifies the outcome as "the most cooling of deadly drugs"); and @altered="true" means the proposition phase is similar to Plutarch's phase but diverges in details (the proposition phase focuses on the scraping process, while Plutarch focuses on what emerges from it). Most proposition phases remain unlinked in Plutarch's simplified version, which condenses the recipe into its essential steps.


These examples show how sequence mapping compares processual theses across sources. The mapping reveals which phases correspond, which remain unlinked, and how different authors describe the same procedures with varying levels of detail, different ordering, or alternative approaches. Map-based or network visualisations can make these differences very evident, showing at a glance which phases link across sources and which remain unique to particular authors.