Data Model‎ > ‎

More Case Studies


1        Introduction

1.1       Personal Names

1.2       Places

1.3       Groups

1.4       Citation Template

1.5       Transcription Anomalies.

1.6       Narrative Reports.

1.7       Dialogue Transcription.


1      Introduction

This paper tackles further case studies in order to test the application of the STEMMA® Data Model to difficult scenarios. They follow on from the case studies already presented under STEMMA Case Studies.

 

Note: This text capitalises entity names like Person and Place to distinguish them from common English usage.

1.1    Personal Names

When we look at the sources of evidence for a person, we nearly always find variations of their personal name. These may be misspellings, recording errors, or transcriptions errors, but they may also be variations that the person accepted themselves, e.g. the use of a middle name, or a diminutive form such as ‘Tony’ instead of ‘Anthony’.

 

In some cases, we may not be certain so how do we handle the variations? How do we distinguish what we believe to be errors from what we believe to be accepted variations?

 

The following case is based on a real person whose middle name began with a ‘K’ but we cannot be certain what it stood for. Susanna Richmond was born out of wedlock in 1827 to Jane Richmond. On her baptism record her middle name was recorded as ‘Kindle’, and on her marriage record her middle name was recorded as ‘Kenyon’. In all other sources only the middle initial was used. One of these is probably wrong but we only have one source supporting each variation.

 

<Person Key=’pSusannaRichmond’>

<Title> Susanna Richmond (1827) </Title>

<Sex> 0 </Sex>

<Names>

<Sequences UntilEvent=’eMarriageSRichmond’>

<Canonical> Susanna K. Richmond </Canonical>

<Sequence>

<Tokens>

<Token> Susanna </Token>

<Token> Susan </Token>

</Tokens>

<Tokens Optional=’1’>

<Token> K<Ucf>[ie]</Ucf>ndle </Token>

<Token> Kenyon </Token>

</Tokens>

<Tokens>

<Token> Richmond </Token>

<Token> Procter </Token>

<Token> Proctor </Token>

</Tokens>

</Sequence>

</Sequences>

<Sequences FromEvent=’eMarriageSRichmond’

Type=’Married’>

<Canonical> Susan K. Wheat </Canonical>

<Sequence>

<Tokens>

<Token> Susan </Token>

</Tokens>

<Tokens Optional=’1’>

<Token> Kenyon </Token>

</Tokens>

<Tokens>

<Token> Wheat </Token>

</Tokens>

</Sequence>

</Sequences>

</Names>

 

<Birth><Eventlet>

<When>

<Date><Value Margin=’1’> 1827 </Value></Date>

<Constraints>

<Constraint BeforeEvent=’eBaptismSRichmond’/>

</Constraints>

</When>

<Type> Birth </Type>

<SubType> Birth </SubType>

<PlaceLnk Key=’wNottm’/>

</Eventlet></Birth>

</Person>

 

<Event Key=’eBaptismSRichmond’>

<Title> Baptism of Susanna Richmond </Title>

<When><Date><Value> 1827-03-11 </Value></Date></When>

<Type> Religious </Type>

<SubType> Baptism </SubType>

<PlaceLnk Key=’wNottmStMary’/>

<SourceLnk Key=’sParishBptSRichmond’>

<PlaceLnk Key=’wMountEastSt’ Key=’wMountESt’>

<Property Name=Name’> Mount East Street </Property>

</PlaceLnk>

<PersonLnk Key=’pSusannaRichmond’>

<Property Name=’Name’>

Susanna K<Ucf>[ie]</Ucf>ndle Richmond

</Property>

<Property Name=’ResidencePlace’  Key=’wMountESt’/>

 

<Text>

Second letter of name looks like an ‘i’ but could possibly be an ‘e’. No father was recorded. Kindle is a rare surname. However, a married Samuel Kindell (Occ: ‘Lace Maker’) also lived on Mount East St at the same time and had a son, Samuel, baptised on 9/12/1827 (the same year).

</Text>

</PersonLnk>

</SourceLnk>

</Event>

 

<Event Key=‘eCensusRichmond1841’>

<Title> Susanna Procter in 1841 census of England </Title>

<When><Date><Value> 1841-06-06 </Value></Date></When>

<Type> Survey </Type>

<SubType> Census </SubType>

<PlaceLnk Key=’wIndChapelSneinton’/>

<SourceLnk Key=’sCensusRichmond1841’>

<PersonLnk Key=’pSusannaRichmond’>

<Property Name=’Name’> Susanna Procter </Property>

<Property Name=’Age’> 14 </Property>

<Property Name=’Occupation’> F.S. </Property>

<Text>

Working for the Place family.

</Text>

</PersonLnk>

</SourceLnk>

</Event>

 

<Event Key=’eMarriageSRichmond’>

<Title> Marriage of Susan Richmond to John Wheat </Title>

<When><Date><Value> 1867-12-25 </Value></Date></When>

<Type> Union </Type>

<SubType> Marriage </SubType>

<PlaceLnk Key=’wIndChapelSneinton’/>

<SourceLnk Key=’sParishMarriageSRichmond’>

<PersonLnk Key=’pSusannaRichmond’>

<Property Name=’Name’> Susan Kenyon Richmond </Property>

<Property Name=’Age’ Value=’?’> Full age </Property>

<Property Name=’Role’> Bride </Property>

</PersonLnk>

<PersonLnk Key=’pThomasProctor’>

<Property Name=’Name’> Thomas Richmond Proctor </Property>

<Property Name=’Role’> Witness </Property>

<Text>

This is Susanna’s half-brother

</Text>

</PersonLnk>

<PersonLnk>

<Property Name=’Name’> Martha Holmes </Property>

<Property Name=’Role’> Witness </Property>

</PersonLnk>

</SourceLnk>

</Event>

 

The <Names> element of the Person entity is effectively a conclusion. In there, we’ve separated the names used before her marriage from the names used thereafter. The actual names recorded in each source case, whether right or wrong, are represented by the Property called ‘Name’ against the corresponding source. Remember that Properties are extracted and summarised items of information from some source, and each Event is backed by one-or-more sources.

 

Each division of the <Names> element is called <Sequences> and serves two main purposes. It defines a number of canonical personal names used for output, and a number of token sequences that can be used for the matching of a name against this Person.

 

Although the code doesn’t present all the available evidence, we would find that she was called by both Susanna and Susan before her marriage. She also used the married surname of her mother (Proctor/Procter) for while, probably until she was old enough to choose herself. On her baptism, her middle name appears as ‘Kindle’ (possibly ‘Kendle’) and on her marriage certificate it appears as ‘Kenyon’. From her marriage onward, she occasionally used the middle initial ‘K.’ but there were no further instances of an explicit middle name. Also, she appears to have abandoned her original name of Susanna in favour of Susan.

 

Given the uncertainty of the middle name, the two canonical names (separated by the marriage event) only specify the middle initial, although the token sequences include the actual source instances.

 

Notice, too, that the Person entity also has a title which doesn’t have to be any of her personal names. In this instance, we elected to pick her birth name and annotate it with her date-of-birth. Another alternative might have been to specify her maiden and married surnames separated by a solidus (‘/’), such as ‘Susanna Richmond/Wheat’. It all depends on the convention you want to adopt as that title field is not constrained.

 

Some small features to take notice of are:

 

  • The Property values can use <Ucf> character sequences to indicate uncertain characters. These can also be used in the tokens of the <Names> element.
  • On her marriage certificate, her age is given simply as ‘Full age’. The relevant <EventLnk> shows how to handle this by recording what was written but without a numeric interpretation. The relevant parts of the STEMMA specification to consult for this are PROPERTY_VALUE and value-notation.
  • There are no sources directly supporting the birth Event. However, we have used an Event-constraint to indicate that it must have come before the baptism Event. See Event Constraints.

 

The importance of this case study is not in prescribing how to handle all cases but in demonstrating the flexibility to handle many different cases. The essential components may be summarised briefly as:

 

  • Recorded names represented verbatim as Properties for each sourced Event.
  • Conclusion names represented as time-dependent data for the Person.
  • Conclusion names consist of separate canonical names and token sequences used for controlling name matching.
  • Canonical names may be given a number of attributes to distinguish the type of name (e.g. familiar, formal, professional, etc) and the form (e.g. language, phonetic, Romanised).
  • Each Person also has a fixed title that can be used in reports or charts for identification. This is unconstrained.

 

1.2    Places

Most researchers are primarily interested in people, and places are merely referenced in relation to those people, e.g. a place of residence. STEMMA was designed to take advantage of strong analogies between Person and Place with the possible goal of it being applied to the study of places, including One-Place Studies.

 

This slightly contrived example gives an illustration of how that analogy works. Instead of Person entities connected to related Events, each supported by one or more sources, we show the same idea with Places.

 

For this illustration, we pick two neighbouring households on a small road called Manning Grove in Nottingham, England.

 

<Dataset Name=’Place_Example’

xmlns:props=’http://mydomain.com/properties’>

 

<ExtendedProperties>

<PlaceProperties>

<PropertyDef Name=’props:HeadHouse’  Type=’PersonRef’/>

</PlaceProperties>

</ExtendedProperties>

 

<Place Key='wManningGrove15'>

<Title> 15 Manning Grove </Title>

<Type> Number </Type>

<PlaceName> 15 </PlaceName>

<ParentPlaceLnk Key='wManningGrove'/>

</Place>

 

<Place Key='wManningGrove17'>

<Title> 17 Manning Grove </Title>

<Type> Number </Type>

<PlaceName> 17 </PlaceName>

<ParentPlaceLnk Key='wManningGrove'/>

</Place>

 

<Place Key='wManningGrove'>

<Title> Manning Grove </Title>

<Type> Street </Type>

<PlaceName> Manning Grove </PlaceName>

<ParentPlaceLnk Key='wNottm'/>

</Place>

 

<Place Key='wNottm'>

<Title> Nottingham </Title>

<Type> Town </Type>

<PlaceName> Nottingham </PlaceName>

<ParentPlaceLnk Key='wNotts'/>

</Place>

 

<Place Key='wNotts'>

<Title> Nottinghamshire </Title>

<Type> County </Type>

<PlaceName> Nottinghamshire </PlaceName>

<ParentPlaceLnk Key='wEngland'/>

</Place>

 

<Event Key='eDirectory1885_Notts'>

<When Value='1885'/>

<Title> Directory of Nottinghamshire, 1885 </Title>

<Type> Survey </Type>

<SubType> Directory </SubType>

<PlaceLnk Key='wNotts'/>

 

<SourceLnk key=’sHistoryGazetteerDirectoryNotts’>

<PlaceLnk Key=’wManningGrove15’>

<Property Name='props:HeadHouse'> Frank Camm </Property>

</PlaceLnk>

<PlaceLnk Key=’wManningGrove17’>

<Property Name='props:HeadHouse'> George F. Nugent </Property>

</PlaceLnk>

</SourceLnk>

</Event>

 

<Event Key='eCensus1891_ManningGr'>

<When Value='1891-04-05'/>

<Title> 1891 census for Manning Grove </Title>

<Type> Survey </Type>

<SubType> Census </SubType>

<PlaceLnk Key='wManningGrove'/>

 

<SourceLnk Key=’sCensusEngWales1891'>

<PlaceLnk Key='wManningGrove15'>

<Property Name='props:HeadHouse'> Fred Smith </Property>

</PlaceLnk>

<PlaceLnk Key='wManningGrove17'>

<Property Name='props:HeadHouse'> Walter Farnsworth </Property>

</PlaceLnk>

</SourceLnk>

</Event>

 

<Event Key='eCensus1901_ManningGr'>

<When Value='1901-03-31'/>

<Title> 1901 census for Manning Grove </Title>

<Type> Survey </Type>

<SubType> Census </SubType>

<PlaceLnk Key='wManningGrove'/>

 

<SourceLnk Key=’sCensusEngWales1901’>

<PlaceLnk Key=’wManningGrove15’>

<Property Name='props:HeadHouse'> Joseph Castle </Property>

</PlaceLnk>

<PlaceLnk Key=’wManningGrove17’>

<Property Name='props:HeadHouse'> John Bullock </Property>

</PlaceLnk>

</SourceLnk>

</Event>

 

<Event Key='eCensus1911_ManningGr'>

<When Value='1911-04-02'/>

<Title> 1881 census for Manning Grove </Title>

<Type> Survey </Type>

<SubType> Census </SubType>

<PlaceLnk Key='wManningGrove'/>

 

<SourceLnk Key='sCensusEngWales1911'>

<PlaceLnk Key='wManningGrove17'>

<Property Name='props:HeadHouse'> W. M. Chapman </Property>

</PlaceLnk>

<PlaceLnk Key='wManningGrove15'>

<Property Name='props:HeadHouse'> G. Rogers </Property>

</PlaceLnk>

</SourceLnk>

</Event>

 

<Source Key='sCensusEngWales1891'>

<Frame>

<CitationLnk Key='cCensusEngWales'>

<Param Name='Series'> RG12 </Param>

<Param Name='Piece'> 2694 </Param>

<Param Name='Folio'> 124 </Param>

</CitationLnk>

</Frame>

<SourceLet Key=’lP19’>

<Frame>

<CitationLnk Key='cCensusEngWales'>

<Param Name='Page'> 19 </Param>

</CitationLnk>

</Frame>

</SourceLet>

<SourceLet Key=’lP20’>

<Frame>

<CitationLnk Key='cCensusEngWales'>

<Param Name='Page'> 20 </Param>

</CitationLnk>

</Frame>

</SourceLet>

</Source>

 

<Source Key='sCensusEngWales1901'>

<Frame>

<CitationLnk Key='cCensusEngWales'>

<Param Name='Series'> RG13 </Param>

<Param Name='Piece'> 3178 </Param>

<Param Name='Folio'> 51 </Param>

<Param Name='Page'> 12 </Param>

</CitationLnk>

</Frame>

</Source>

 

<Source Key='sCensusEngWales1911'>

<Frame>

<CitationLnk Key='cCensusEngWales1911'>

<Param Name='Series'> RG14 </Param>

<Param Name='Piece'> 20565</Param>

<Param Name='RD'> 430 </Param>

<Param Name='SD'> 3 </Param>

<Param Name='ED'> 25 </Param>

</CitationLnk>

</Frame>

<SourceLet Key=’lS37’>

<Frame>

<CitationLnk Key='cCensusEngWales1911'>

<Param Name='Schedule'> 37 </Param>

</CitationLnk>

</Frame>

</SourceLet>

<SourceLet Key=’lS38’>

<Frame>

<CitationLnk Key='cCensusEngWales1911'>

<Param Name='Schedule'> 38 </Param>

</CitationLnk>

</Frame>

</SourceLet>

</Source>

 

So what we’ve done here is to register a new place-property called HeadHouse to represent the head of the household. The data type of this Property is PersonRef and so any of those Property values could be given a link to an actual Person entity. For instance:

 

<Property Name='props:HeadHouse' Key=’pJosephCastle’> Joseph Castle </Property>

 

This would be useful if you were doing a little bit of the history of the people in each household. Alternatively, if that historical analysis wanted to assemble profiles of the various subject entities, including their relationships, before making a final conclusional association, then this could be done in a more flexible way using the Source entity:

 

<Source Key='sCensusEngWales1901'>

… Frame etc …

<ProtoPerson DetKey=‘dpJosephCastle’ Key=‘pJosephCastle’>

<Link DetLnk=’dtJosephCastle’ Value=’Joseph Castle’ Type=’Source’>

<Text>personal name</Text>

</Link>

<Link DetLnk=’dtJosephAge’ Value=’24’ Type=’Source’>

<Text>age</Text>

</Link>

</ProtoPerson>

 

<ProtoPlace Key=‘dwManningGrove15’>

<Link DetLnk=’dtStreet’ Value=’Manning Grove’>

<Text>street</Text>

</Link>

<Link DetLnk=’dtNumber’ Value=’15’>

<Text>house number</Text>

</Link>

<Link DetLnk=’dtHead’  Value=’Head’ Type=’Source’>

<Text>role</Text>

</Link>

<Link DetLnk=’dpJosephCastle’ Type=’Inference’>

<Text>head of household</Text>

</Link>

</ProtoPlace>

</Source>

 

This builds up a profile of the place and of the man called Joseph Castle. It then associated him with the place as the head of that household. A network, of extracted information and logic can be used to make the connection with the conclusion Place (wManningGrove) and the conclusion Person (pJosephCastle) in a form of Link Analysis.

 

By contrast, the Property name-value pairs represent extracted and summarised data in a normalised (computer-readable) way.

 

<SourceLnk Key=’sCensusEngWales1901’>

<PlaceLnk Key=’wManningGrove15’>

<Property Name='props:HeadHouse'> Joseph Castle </Property>

</PlaceLnk>

… etc …

</SourceLnk>

 

1.3    Groups

As of V2.2, STEMMA gives the same top-level support to Groups as it always did to Persons and Places. The example chosen here is based on a reference to a military regiment in a blog post at: A Grave Too Far.

 

The reference concerned a British cavalry regiment known as the 14th/20th Hussars and its movement to Risalpur in British India[1]. Although important to the research article, there was no mention of any specific persons in the sources supporting that event. The article also made a note about the history of the regiment’s name which we will also incorporate in this example.

 

 

<Group Key=’g1420Hussars’>

<Title> 14th/20th Hussars) </Title>

<Type> Military </Type> </SubType> Travel </SubType>

<Names>

<Sequences Until=’1936’>

<Canonical> 14th/20th Hussars </Canonical>

</Sequences>

<Sequences From=’1936’>

<Canonical> 14th/20th King’s Hussars </Canonical>

</Sequences>

</Names>

 

<Creation>

<Eventlet>

<Type> Military </Type>

<When Value=’1922’/>

</Eventlet>

<JoinFrom Key=’g14KingsHussars’/>

<JoinFrom Key=’g20Hussars’/>

 

<Text>Created through the merger of the 14th King's Hussars and the 20th Hussars in 1922. The honorific "King's" was added back into the title in 1936 (<Link Url=’https://en.wikipedia.org/wiki/14th/20th_King's_Hussars’> 14th/20th Hussars</Link>)</Text>

</Creation>

 

<Eventlet>

<When Value=’1934-01-09’/>

<Type> Military </Type> <SubType> Travel </SubType/>

<SourceLnk Key=’sTheTimes1934’/>

</Eventlet>

</Group>

 

<Citation Key=’cTheTimes1934’>

URI “http://stemma.parallaxview.co/source-type/newspaper”

<Params>

<Param Name=’Title’> Cavalry Change at Risalpur </Param>

<Param Name=’Paper’> The Times </Param>

<Param Name=’Date’> 1934-01-09 </Param>

<Param Name=’Page’> 15 </Param>

</Params>

 

<Text>

<ts>

The 14th/20th Hussars from the Cairo Cavalry Brigade arrive at Karachi today in the transport Nevasa. The regiment will go by rail to Risalpur, and take up duty in the 1st Cavalry Brigade on Thursday in relief of the 15th The King’s Royal Hussars who leave Risalpur on Friday to embark for home.

</ts>

</Text>

</Citation>

 

<Source Key=’cTheTimes1934’>

<Frame>

<CitationLnk Key=’cTheTimes1934’/>

</Frame>

</Source>

 

 

Some points to note here: The transcription of the newspaper report for the troop movements was placed in the Citation entity. If there were also a Resource entity representing a scan then it could have been placed in there. This is discussed further in Handling Transcriptions. Also, the names are presented only in their Canonical form. In the case of a single such form then the relevant Tokens can be deduced by a simple tokenisation operation when the data is loaded.

 

1.4    Citation Template

Although STEMMA’s citation-template support is simplistic, and largely a by-product of supporting hand-crafted citations, this example illustrates a classic use of it.

 

The following entity defines an Abstract Citation from which many separate book Citations can be derived. It includes the source-type, parameters, and a default display-format to generate an English reference-note citation.

 

<Citation Key=’cBook’ Abstract=’1’>

<URI> http://stemma .parallaxview.co/source-type/book/ </URI>

<Params>

<Param Name=’Author’/>

<Param Name=’Title’/>

<Param Name=’Publisher’/>

<Param Name=’Date’ Type=’Date’/>

<Param Name=’Pages’/>

</Params>

<DisplayFormat Mode=’RefNote’>

<Text Language=’eng’>

<Subs>${Author},<i>${Title}</i> (${Publisher}, <DateRef Value=’${Date}’/>), p.${Pages}</Subs>

</Text>

</DisplayFormat>

</Citation>

 

Although this citation-template is designed for Language=’eng’, an alternative of Locale=’en-GB’ or =’en-US’ could be specified to differentiate the way quotes and punctuation are handled together.

 

The next Citation entity represents a specific book, although the page numbers are currently unspecified.

 

<Citation Key=’cOldNottm’>

<Title>Old Nottingham Notes</Title>

<BaseCitationLnk Key=’cBook’/>

<Params>

<Param Name=’Author’>James Granger</Param>

<Param Name=’Title’>OLD NOTTINGHAM : Its Streets, People, etc</Param>

<Param Name=’Publisher’>Nottingham Daily Express Office</Param>

<Param Name=’Date’>1904</Param>

<Param Name=’Pages’/>

</Params>

 

<Text>

Reprinted from the Nottingham Daily Express, October 3rd, 1903 – July 9th, 1904

</Text>

</Citation>

 

A corresponding usage of this entity might appear as:

 

<CitationRef Key=’cOldNottm’>

<Param Name=’Pages’>46-48</Param>

</CitationRef>

 

This might result in a reference-note of the form:

 

3 James Granger, OLD NOTTINGHAM : Its Streets, People, etc (Nottingham Daily Express Office, 1904), p.46-48.

 

The use of <CitationLnk> or <CitationRef> for different parts of a given source allows a hosting program to determine whether a short or long reference-note is applicable when not explicitly indicated.

 

Although the use of the <DateRef> element doesn’t appear to contribute much here — it still appears as it was entered: “1904” — it would have helped had the date included more detail. For instance, a STEMMA date of “1904-03” would have been formatted as “March 1904”.

 

1.5    Transcription Anomalies

The following transcript of a Civil Service evidence-of-age declaration first appeared in A Copyright Casualty — Part III. However, much of the original detail was “flattened”, or removed, so that it was readable in that article, and also so that the remaining detail could be represented with simple textual annotation. For instance: ^inserted-text^ and <deleted-text>.

 

A diplomatic transcription (“diplomatic” as in “document”) attempts to capture the essence a document (usually a manuscript one) with a system of editorial annotation, or mark-up, and is therefore distinct from any type of image copy or photographic facsimile.[2] Typed mark-up is considerably more limited than the original handwritten mark-up because it has less freedom with its choice of symbols and mechanisms, and the term semi-diplomatic may be used to indicate that not all features have been represented. Hence, the aforementioned article had presented a simple typed mark-up to display a semi-diplomatic transcription.

 

Both typed and handwritten mark-up are designed to be humanly-readable, but a mark-up language — where the meta-data has its own grammar allowing it to be distinguished from the underlying text — is designed to be computer-readable. Such schemes can capture many more of the original details, and with considerably less ambiguity. The way that the text is displayed to an end-user is then merely dependent upon the capabilities of the display medium and the sophistication of the display software. Both STEMMA and TEI employ mark-up languages.

"Civil service evidence of age", database with images, Findmypast (www.findmypast.co.uk : accessed 23 Oct 2015), entry for William Ashbee born 1836; Surviving records deposited at the Society of Genealogists (SoG) by the Civil Service Commission (CSC). Image courtesy of Findmypast.

 

<Resource Key=’rEvidenceOfAge’>

<Text>

<ms>

<Anom Mode=’Marginalia’ Posn=’T’>

ASHBEE, William

</Anom>

<Anom Mode=’Marginalia’ Posn=’T’>

b 29 Nov 1834

</Anom>

<Anom Mode=’Marginalia’ Posn=’L’>

<ts>Here insert name of Declarant</ts>

</Anom>

<ts>I</ts> William Ashbee of Tetbury in the County<br/>

of Gloucester Baker<br/>

<ts>do solemnly and sincerely declare that</ts> I have looked at the<br/>

paper writing hereunto annexed and marked<br/>

with the letter "A" purporting to be an Extract<br/>

from the Register Book of Baptisms in the<br/>

parish of Westonbirt in the County of Gloucester<br/>

and that William, the son of William and<br/>

Anne "Ashby" therein referred to, is the same<br/>

person as William Ashbee my Son<br/>

who was born on the twenty ninth day<br/>

of November one thousand eight hundred<br/>

and thirty six, his Birth was never registered<br/>

but <s>from</s> <Anom Mode=’Marginalia’ Posn=’L’>RCP<Text> Solicitor’s initials </Text></Anom>

an entry made by my wife<br/>

in a family Bible belonging to me, and<br/>

<s>which I now</s> <Anom Mode=’Marginalia’ Posn=’L’>RCP</Anom> produced at the time of my<br/>

<Anom>making this Declaration</Anom> <s>Birth</s> <Anom Mode=’Marginalia’ Posn=’L’>RCP</Anom> enabled me to speak confidently<br/>

to the fact

</ms>

</Text>

</Resource>

 

As well as distinguishing the boilerplate typescript sections from the manuscript sections, this records the insertions and deletions. It also indicates that those corrections were all initialled with “RCP”: those of the solicitor taking the statement. He was actually Robert C. Paul, and the 1861 census (Piece 1780, Folio 44, Page 35) showed that he was born c1805 and worked on Long Street, Tetbury, not far from the William Ashbee making the statement.

 

1.6    Narrative Reports

The concept of Narrative Reports was described in Our Days of Future Passed — Part II. Several examples are published on my Parallax View blog, but one in particular (Jesson Lesson) was selected to showcase the changes in STEMMA V4.1, the majority of which were related to narrative mark-up and transcription.

 

This 5000-word research article includes precise layout, transcribed extracts, tabulation, endnotes and tablenotes, and hyperlinked images. Its 47 endnotes include all of the cases: reference note citations, discursive notes, analytical commentary, and multi-source references — the handling of which was outlined previously at Cite Seeing — but also include examples of conflated citations where details of multiple people are placed in a single note for readability.

 

The STEMMA representation accurately captures the structure and content of that article, and a transformation to HTML mirrors the presentation achieved on the blog. The XML for this fully-worked example is available at www.parallaxview.co/familyhistorydata/downloads/JessonLesson.xml.

 

Another worked example may be found at www.parallaxview.co/familyhistorydata/downloads/MoreOnGeorgeHearson.xml. This is a STEMMA version of the earlier blog-post, More on George Hearson, and is noteworthy because it contains an involved example of a citation chain together with commentary and substituted text for missing citation-element information. These relate to three images of a handbill from 1832, and the example deals with cases where a hereinafter-cited-as term is used in subsequent references with elided citation-element information. This particular citation scenario was the subject of a Q&A on the Evidence Explained Web site at Chicken-and-Egg.

 

1.7    Dialogue Transcription

The following example illustrates STEMMA’s support for audio transcription using a simple dialogue between two individuals that includes overlapping audio contributions, pauses, and intonations.

 

STEMMA’s general approach to transcription separates structure and content from presentation, and applies both to textual and audio cases. One the transcription is marked-up to identify the structure and content, then any specific details pf presentation or style including intonation, different voices, different hands, different colours, etc. can be provided in narrative.

 

<Resource Key=’rRecordings’ Abstract=’1’>

<Title> Voice recording: ${File} </Title>

<Type> Recording </Type>

<URL ContentType=’audio/mpeg’> file:myrecordings/sound/{$File}.mp3 </URL>

<Params>

<Param Name=’File’/>

</Params>

</Resource>

 

<Text>

<Text Class=’Legend’>

Example dialogue between two people over a piece of music.<br/>

Participants:<br/>

<indent/>P1 - First person<br/>

<indent/>P2 - Second person<br/>

<indent/>Music – Music from a record player<br/>

Audio schemes:<br/>

<indent/>Softly – Spoken in a hushed tone<br/>

</Text>

 

<ResourceRef Mode=’SynchAudio’>

<Param Name=’File’> Conv_2017_03_20 </Param>

</ResourceRef>

 

<voice id='P1'>

So do you have a favourite piece of film music?

</voice>

 

<voice id='P2'>

Yes, and it's not a well-known piece. It was written by Paul Ferris and used as the theme to the 1968 film Witchfinder General, starring Vincent Price.

</voice>

 

<voice id='P1'>

Oh, I remember that. Didn't it have a synthesizer score?

</voice>

 

<voice id='P2'>

Ah, that was the US version. The UK version has this beautiful orchestral piece that seemed incongruous in a film about the violent and tragic plight of so-called witches at the hands of a sadist. Listen to this…

</voice>

 

He then placed a record on his turntable and began lowering the needle...

 

<voice bg=’1’ id=’Music’>

<time stamp='00:41'/>

 

<voice id='P2' scheme='Softly'>

The first two minutes is the romanza, <Anom Mode='Gesture' Descr=’finger over pursed lips’>and would have been used during the credits.</Anom>

<Anom Mode=’Pause’ Dur=’01:49’/>

The main score is gentle and melodic, mirroring the rhythm of an early-morning horse-ride across the <Alt Mode=’Inline’ Value=’East Anglia, England’>Suffolk</Alt> fields.

<Anom Mode=’Pause’ Dur=’02:14’/>

There's a profound sadness in the music that befits remembrance of those poor women.

</voice>

 

<time stamp='07:05'/>

</voice>

</Text>

 

While person P2 is talking over the music, there are long pauses between his sentences. The timestamps (<time> element) at the start and end of that overlap link to specific points in a pre-recorded MP3 file. It might be unusual to have both timestamps and pauses since they would be hard to keep in step. Timestamps are only relevant where there is an associated recording (see ‘SynchAudio’), and pauses are generally just approximations.

 

The identification of the parties P1 and P2 in this example doesn’t really do them justice. They could have easily been linked to specific Person (or Contact) entities in the data using <PersonRef> mark-up.



® STEMMA is a registered trademark of Tony Proctor.

[1] "Cavalry Change at Risalpur", The Times (Tuesday 9 Jan 1934): p.15.

[2] A discussion on the difference between diplomatic transcription and typographic facsimile may be found in: Mary-Jo Kline and Susan Holbrook Perdue, Guide to Documentary Editing, 3rd edition, chapter 5, section IV (http://gde.upress.virginia.edu/05-gde.html#h2.4 : accessed 16 Oct 2013).

ą
Tony Proctor,
21 Nov 2015, 05:38