Around the 1230s, Hugh of Saint-Cher and a team of Dominican scholars produced the first Bible concordance, working through the Latin Vulgate and recording where its words appeared. Because modern verse numbers did not yet exist, the compilers divided chapters into smaller lettered sections so readers could locate an occurrence.
Printing made the method far more useful. John Marbeck published the first complete concordance to the English Bible in 1550. Alexander Cruden's Complete Concordance followed in 1737 and became one of the best-known English Bible reference works. The basic structure remained familiar for centuries: choose a word, then receive a list of passages where it occurs.
Latter-day Saints soon adopted the same approach for their own scriptures. In 1900, George Reynolds published A Complete Concordance to the Book of Mormon in Salt Lake City. The 852-page volume was the product of roughly twenty years of work and classified the words of the Book of Mormon so readers could trace passages across the text. It was an enormous manual indexing project, built for the same basic purpose as the Bible concordances that came before it.
Moving into the modern age, computers changed the scale. Roberto Busa's Index Thomisticus helped pioneer machine-generated concordances in the middle of the twentieth century, and in 1957 a UNIVAC computer was used to compile a concordance to the Revised Standard Version of the Bible. Work that once required years of manual indexing could increasingly be done by machines.
The Church later folded concordance-like tools directly into its standard works. Its 1979 Latter-day Saint edition of the King James Bible included the Topical Guide, Bible Dictionary, and extensive cross-references, and the 1981 triple combination added improved footnotes and an index for the Book of Mormon, Doctrine and Covenants, and Pearl of Great Price. The result was a much more interconnected set of printed scriptures, with one passage pointing readers toward related passages elsewhere in the canon.
But even a digital concordance usually behaves like an index, which many people find helpful, but I wanted to take it one step further.
Lucifer’s Lantern has developed a program called Concord which helps readers visually see the relationships and cross-references between texts. The intent is to allow a reader to see the scriptures in a way that a simple index can’t.
The Church of Jesus Christ of Latter-day Saints identifies four volumes as its standard works: the Bible, the Book of Mormon, the Doctrine and Covenants, and the Pearl of Great Price. Readers usually encounter them one book, chapter, and verse at a time.
Concord has the entire standard works uploaded into its corpus, representing every verse passages as a node. Connections between those nodes can show repeated phrases, related concepts, quotations, paraphrases, shared images, and larger narrative or structural similarities. A passage in Isaiah can be viewed beside related material in the Book of Mormon. A cluster can show several passages using the same image in different settings. A longer sequence can reveal that two chapters follow a similar narrative progression even when their wording is not especially close.
The visual approach changes the question from "Where else does this word appear?" to "What is this passage connected to?"
That matters because some relationships are easy to miss when viewing a massive list of text.. By visualizing the connections, the reader can have a different experience. Several passages that repeatedly connect to one another begin to form a cluster. Links across books are visible at the same time as links within a book. The reader can move through those relationships rather than working through pages of search results one at a time.
At the center of Concord is a verse-level comparison of the LDS canon. Each tick mark on the center line represents a single verse. Each arc is a connector between two verses. Each book can be compared internally with its own text to identify internal references and similarities.
Or against another source, like this comparison between the Old Testament and the Book of Mormon
Focal mode allows multiple texts to be compared against one focal text as well. Like this example of The Old and New Testaments being compared against the Book of Mormon but not to each other. This allows for more nuanced studies.
When a user first loads the page, the initial connections displayed consist of high confidence connections based on existing research. These are the highest confidence connections. By clicking "Find Connections" the user triggers the server to attempt to identify connections on any of the visible texts. Each run reinforces its internal model.
The visual map gives those findings a physical shape. Selecting a node shows the passage itself along with the passages Concord has connected to it. Readers can follow a line, inspect the texts side by side, and decide whether the relationship is meaningful. All votes and user notes go back to the server (anonymously) and help aid Concord in its continuous training. (This feels like a good time to let you know that concord is NOT an LLM and could instead be considered to be a specialized Natural Language Processing Intertextual Analysis Engine)
In addition to the graph, each full text can be explored in a simple reading mode.
The user has an option to run a report, which generates a PDF ready presentation of every connection currently visible on the graph. This can allow for deeper study of the corpus texts. Reports improve with time as the program continues to train on data gathered by the community, through its automated training schedule, and data that I personally identify and feed through the backend. (The tool itself does not scrape the web)
Concord does not treat every match as equal. Results are weighed by confidence in the comparison and the complexity of the similarities for example, A long, multi-verse relationship with several reinforcing features should carry more weight than a short phrase that happens to occur twice. Stronger and more complex findings are brought forward while weaker or simpler similarities remain available farther down the report.
A line between two passages is not proof that one author copied another. Similarity can come from direct borrowing, a shared source, a common religious vocabulary, a familiar narrative pattern, coincidence, or something else entirely.
The point is to make possible connections easier to see and easier to inspect. Once a connection is surfaced, the ordinary work of research still follows. It is still your responsibility to read passages need to be read in context. Dates matter. Sources matter. Access matters. A visual connection gives the reader somewhere to look.
For centuries, concordances helped readers find their way back into a text. Concord keeps that basic purpose, but adds another dimension.
Printing made the method far more useful. John Marbeck published the first complete concordance to the English Bible in 1550. Alexander Cruden's Complete Concordance followed in 1737 and became one of the best-known English Bible reference works. The basic structure remained familiar for centuries: choose a word, then receive a list of passages where it occurs.
Latter-day Saints soon adopted the same approach for their own scriptures. In 1900, George Reynolds published A Complete Concordance to the Book of Mormon in Salt Lake City. The 852-page volume was the product of roughly twenty years of work and classified the words of the Book of Mormon so readers could trace passages across the text. It was an enormous manual indexing project, built for the same basic purpose as the Bible concordances that came before it.
Moving into the modern age, computers changed the scale. Roberto Busa's Index Thomisticus helped pioneer machine-generated concordances in the middle of the twentieth century, and in 1957 a UNIVAC computer was used to compile a concordance to the Revised Standard Version of the Bible. Work that once required years of manual indexing could increasingly be done by machines.
The Church later folded concordance-like tools directly into its standard works. Its 1979 Latter-day Saint edition of the King James Bible included the Topical Guide, Bible Dictionary, and extensive cross-references, and the 1981 triple combination added improved footnotes and an index for the Book of Mormon, Doctrine and Covenants, and Pearl of Great Price. The result was a much more interconnected set of printed scriptures, with one passage pointing readers toward related passages elsewhere in the canon.
But even a digital concordance usually behaves like an index, which many people find helpful, but I wanted to take it one step further.
Lucifer’s Lantern has developed a program called Concord which helps readers visually see the relationships and cross-references between texts. The intent is to allow a reader to see the scriptures in a way that a simple index can’t.
Seeing the canon
The Church of Jesus Christ of Latter-day Saints identifies four volumes as its standard works: the Bible, the Book of Mormon, the Doctrine and Covenants, and the Pearl of Great Price. Readers usually encounter them one book, chapter, and verse at a time.
Concord has the entire standard works uploaded into its corpus, representing every verse passages as a node. Connections between those nodes can show repeated phrases, related concepts, quotations, paraphrases, shared images, and larger narrative or structural similarities. A passage in Isaiah can be viewed beside related material in the Book of Mormon. A cluster can show several passages using the same image in different settings. A longer sequence can reveal that two chapters follow a similar narrative progression even when their wording is not especially close.
The visual approach changes the question from "Where else does this word appear?" to "What is this passage connected to?"
Feel free to download any image, or right click and "open image in new tab" so you can zoom in
That matters because some relationships are easy to miss when viewing a massive list of text.. By visualizing the connections, the reader can have a different experience. Several passages that repeatedly connect to one another begin to form a cluster. Links across books are visible at the same time as links within a book. The reader can move through those relationships rather than working through pages of search results one at a time.
How Concord works
At the center of Concord is a verse-level comparison of the LDS canon. Each tick mark on the center line represents a single verse. Each arc is a connector between two verses. Each book can be compared internally with its own text to identify internal references and similarities.
Or against another source, like this comparison between the Old Testament and the Book of Mormon
When a user first loads the page, the initial connections displayed consist of high confidence connections based on existing research. These are the highest confidence connections. By clicking "Find Connections" the user triggers the server to attempt to identify connections on any of the visible texts. Each run reinforces its internal model.
The visual map gives those findings a physical shape. Selecting a node shows the passage itself along with the passages Concord has connected to it. Readers can follow a line, inspect the texts side by side, and decide whether the relationship is meaningful. All votes and user notes go back to the server (anonymously) and help aid Concord in its continuous training. (This feels like a good time to let you know that concord is NOT an LLM and could instead be considered to be a specialized Natural Language Processing Intertextual Analysis Engine)
In addition to the graph, each full text can be explored in a simple reading mode.
Reports
The user has an option to run a report, which generates a PDF ready presentation of every connection currently visible on the graph. This can allow for deeper study of the corpus texts. Reports improve with time as the program continues to train on data gathered by the community, through its automated training schedule, and data that I personally identify and feed through the backend. (The tool itself does not scrape the web)
Concord does not treat every match as equal. Results are weighed by confidence in the comparison and the complexity of the similarities for example, A long, multi-verse relationship with several reinforcing features should carry more weight than a short phrase that happens to occur twice. Stronger and more complex findings are brought forward while weaker or simpler similarities remain available farther down the report.
Built for discovery
A line between two passages is not proof that one author copied another. Similarity can come from direct borrowing, a shared source, a common religious vocabulary, a familiar narrative pattern, coincidence, or something else entirely.
The point is to make possible connections easier to see and easier to inspect. Once a connection is surfaced, the ordinary work of research still follows. It is still your responsibility to read passages need to be read in context. Dates matter. Sources matter. Access matters. A visual connection gives the reader somewhere to look.
For centuries, concordances helped readers find their way back into a text. Concord keeps that basic purpose, but adds another dimension.
Use concord by clicking the image below (note that being on desktop is a better experience)
Comments
Post a Comment