Brian W. Davidson

sharing things I enjoy

My favorite feature of Logos Bible Software is the Ancient Literature Dataset. There is nothing like it in any other Bible software. Logos uses this dataset to provide users with an extensive list of hyperlinked cross-references between scripture and a host of other extra-biblical literature.

Users have access to this dataset via the Passage Guide or by opening a separate Ancient Literature guide. Here is how the dataset documentation describes the feature:

The Ancient Literature Guide Section is a curated and classified index of ancient material relevant to Bible passages with the goal of bridging the gap between canonical material and other ancient literature.

The dataset provides the most extensive and useful cross-referencing system I have ever seen. Here are the bodies of literature to which the dataset provides cross-references:

  • Ancient Near-Eastern Material (Ugaritic material, Context of Scripture, Ancient Near-Eastern Texts, Amarna Letters, etc.)
  • Apostolic Fathers
  • Church Fathers
  • Dead Sea Scrolls (sectarian material)
  • Judaica (Babylonian Talmud, Jerusalem Talmud, Mishnah, Mekhilta, etc.)
  • Old Testament Pseudepigrapha
  • Nag Hammadi Codices
  • New Testament Apocrypha
  • Works of Josephus
  • Works of Philo

References to this material are categorized by various types of textual relationships:

  • Citation
  • Quotation
  • Allusion
  • Echo
  • Historical
  • Topical
  • Lexical
The Ancient Literature Guide displayed in a custom guide

The dataset was created by Rick Brannan and Ken Penner. Rick Brannan is the data engineer at Faithlife who was responsible for aggregating the references, and Ken Penner (St. Francis University) worked as a consultant with respect to the reference classification and terminology.

In this post, Rick Brannan has been kind enough to answer some questions about their work.

Rick Brannan, Faithlife data engineer & co-creator of the Ancient Literature Dataset

BWD: Rick, how did the idea of creating this dataset arise?

RB: I think the first edition of the Ancient Literature Dataset was released with Logos 6. And I really don’t remember what got us going down this road. The key to creating the dataset, though, was realizing that technical commentaries frequently reference ancient writings, and editors of ancient writings frequently reference the Bible.

BWD: That is fascinating! I had no idea commentaries were so important here. I saw this mentioned in the dataset documentation, but I just assumed they would have been more tertiary. It makes sense, though, because without the ancient literature feature, commentaries would be the first place I turn to find connections between scripture and extra-biblical literature. Do you have any sense of which commentary sets produced the most references?

RB: That’s an interesting question because references don’t really come from any single resource or commentary. The process sort of uses the associations it builds from commentaries and editions as votes toward inclusion of a reference in the dataset. That said, if I recall correctly, the Anchor Yale Bible Commentary and the International Critical Commentary, as sets, have larger amounts of references than others. However, commentaries from ancient Christian authors (e.g. Fathers of the Church, Ante-Nicene Fathers, Ancient Christian Writers, etc.) tend to have higher proportions of references as well.

BWD: When you say “references don’t really come from any single resource,” do you mean that you can’t see which resources the references are being pulled from via the automated process?

RB: What I mean by that is that we’re not simply building a list of references that happen to occur in a particular edition of an ancient writing and aggregating them. Consider Polycarp’s letter to the Philippians in the Apostolic Fathers. His writing is infused with phrasing from the New Testament. Yes, there are some outright quotations and certain allusions, but there are a large number of turns of phrase that can be profitably compared to the New Testament. Different translations and editions bring different instances of these phenomena to light. If you build associations across multiple editions and translations (and commentaries), some of these appear stronger with many attestations, others appear weaker with few and even single attestations.

My point here is that if seven different translations or editions note that Poly. Phil. 1.1’s “I rejoice with you greatly in our Lord Jesus Christ” has similarity with Philippians 4:10, which “single resource” does that association come from? Add in licensing factors (a user may only have three of those seven resources available in their library) and it gets complicated quickly.

This is why I stress that we’re building associations between a writing and other material related to the writing, preferring the more strongly attested associations and discounting (and even not including) the less frequently attested associations.

BWD: I see. That explanation is very helpful. I was thinking that the references were gathered both manually and through an automated process.

RB: The references were all gathered through automated processes; the references for some categories (e.g. Apostolic Fathers, Philo, Josephus) were reviewed and classified manually. Other categories (e.g. Church Fathers, Ancient Near-Eastern material) were categorized automatically because of their nature or size of the data.

The process to gather references essentially builds up associations of canonical reference to non-canonical reference through evaluating technical commentaries of canonical material as well as editions, commentaries, and translations of non-canonical texts. These associations are then evaluated and some proportion of them bubble up to the top as most relevant or most frequent. These that “bubble up” are the references that are then further curated (Apostolic Fathers, OT Pseudepigrapha, DSS Sectarian, Josephus, Philo) or categorized in an automated fashion.

The automated categorization is either specified for the whole category (so, Ancient Near-Eastern texts all have a “Topical” relation) or it attempts to deduce level of quotation through evaluating the original languages of the associated references for similarities or for evaluating translations for signals that quotation is happening (material in single or double quotes with proximate footnote indicators, for instance).

BWD: In a recent post, I took a look at the type of data provided for Genesis 1:1. I counted 209 links just to the church fathers, which to me is absolutely amazing. I don’t know of any cross-referencing system that is this extensive. Do you know of any parallels, print or digital?

RB: Yeah, I think everyone makes a comment about Genesis 1:1 at one point or another. Genesis 1:26-27 also has a lot of interaction as well. In print, the material I know of is the Biblia Patristica project, seven print volumes put together I believe in the 1970s that involved an amazing amount of work. As I understand it, this material formed the original basis of the BIBLindex site, which is (for material from ancient Christian authors) the best available parallel. (More information about the development of BIBLindex is here.) I don’t know of a parallel that combines this material, plus the other material included in the Ancient Literature Dataset (Nag Hammadi, Dead Sea Scrolls Sectarian, Josephus, Philo, Judaica, Ancient Near-Eastern Literature).

BWD: As you worked on this dataset and continue to add to it, who is it that you envision as ideal users of the Ancient Literature Guide? What type of students of scripture do you think it can help most? I know that’s a broad question, but I wonder what examples come to your mind most readily.

RB: One of my focuses in the past years here at Faithlife is in trying to get information that is available deep within different resources more easily available for users; the Ancient Literature Guide was probably one of the first efforts in this direction. I don’t know that I have an ideal user envisioned beyond wanting to help that person who knows the information they’re looking for is there but doesn’t know how to go about finding it. My hope is that guides like the Ancient Literature Guide give them a decent shovel to use in digging for the information they are interested in.

BWD: If people want to make the most of the Ancient Literature Guide — that is, they want to be able to open the links found in the guide to ancient authors in full text — what are the top two or three resources they might want to consider purchasing?

RB: This really depends on what sort of material they’re interested in. If they’re interested in Judaica, then the Neusner collection is probably the best single collection, albeit expensive. The alternative would be to make sure one has editions of the Talmud and the Misnhah, I think. For Ancient Near-Eastern material, ensuring one has access to Ancient Near-Eastern Texts (ANET), Context of Scripture (COS), Amarna Letters, and similar aggregations of such material is probably best. For the Old Testament Pseudepigrapha, the best would be Charlesworth’s edition of the material. For the Dead Sea Scrolls Sectarian material, Garcia and Martinez’ two-volume Dead Sea Scrolls: Study Edition covers most of it. For everything else (Josephus, Philo, ancient Christian authors), minimum, I’d think, would be an edition of Josephus, an edition of Philo, an edition of the Apostolic Fathers, and the 37-volume Ante-Nicene, Nicene, and Post-Nicene Fathers. Supplement that with the Fathers of the Church and the Ancient Christian Writers series.

Alternatively or additionally, one could focus on the more academically focused commentary series (Anchor Bible Commentary, International Critical Commentary, Hermeneia) as many times these discuss the associations the Ancient Literature Dataset highlights.

BWD: Rick, one final question: What do people need to know about the Ancient Literature guide? If you were to get twenty seconds to tell people why this feature is worth checking out, what would you say?

RB: Because I’m not in sales, I’d start with a disclaimer: This is an index to material. It doesn’t include translations from the material and may point to things that aren’t in your library. That’s actually one of its strengths, because it can point you to material that you may not otherwise know about that could be profitable to your work.

Next, I’d say to just take time and look at things. Sometimes there are associations that are silver bullets that are obvious and amazing. Other times, you need to sift carefully through material because you’re looking for a needle in a haystack. Give the material your time and serious attention, and it will repay your effort. 

Thanks, Brian, for asking these questions about the Ancient Literature Guide! This data has always been fun to work on, and it is some of the material I turn to most often in my personal study. Thank you for highlighting it and sharing with your readers.

BWD: Rick, thank you so much for your time. It is fascinating to hear more about what’s behind this fantastic feature.

For more information about what it looks like to use the Ancient Literature Dataset/Guide, check out this post.