at all?” “What is the rough context?” “How is the older scholarship citing the page?” “Is the passage easy to consult without specialized access?” Use a critical edition when you need to answer questions such as: “What is the best attested reading?” “Is the work authentic?” “Which manuscripts matter?” “Does this catena preserve a fragment or an altered excerpt?” “Can I publish, translate, or argue from the exact wording?” There are also digital quality issues. Public-domain scans are invaluable, but OCR can be noisy, Greek can be particularly fragile, and facsimile databases are not identical to newly edited texts. Even HathiTrust notes recurring quality problems in OCR and page images, and the open Patrologia Graeca experiments are explicitly OCR-generated raw data, not scholarly editions. A good rule of thumb is this. Free sources are usually sufficient for personal study, sermon preparation, classroom illustration, and initial bibliography-building. Scientific work requires a critical edition once the passage becomes an argument rather than just a lead. The scale of the corpus and the 17,000-books claim The figure 17,000 books is not a stable or well-defined number for “patristic and medieval biblical exegesis” as a whole. I did not find a clear authoritative standard that uses 17,000 as the accepted size of this field. What I did find is that different reference systems count different things: authors, works, commentaries, manuscript witnesses, or bibliographic records. Several anchor points show why the number is unstable: Repertorium Biblicum Medii Aevi alone, for Latin Bible commentaries up to 1500, records around 12,000 numbers and almost 24,000 commentaries. If someone means “medieval Latin commentary records,” then 17,000 is plausible but actually conservative. GlossaOrdinaria.net already lists well over 2,500 manuscripts of the Glossa Ordinaria alone. That is only one work-family, but it shows how large the manuscript side of the tradition becomes very quickly. BiblIndex contains 270,000 entries of biblical references, but those are verse-level references, not books or works. The TLG printed Canon covers 4,500 authors and 20,000 works for Greek literature broadly, not specifically biblical commentaries. So “17,000” could sound plausible if one confuses the whole Greek literary canon with biblical exegesis, but that would be a category error . CPL has 2,348 entries for Latin Christian texts up to Bede. Even allowing for Greek material in CPG, that still means that “17,000 patristic authors” is clearly unrealistic, and “17,000 original patristic Bible commentaries” is also far too high. One unrelated but real 17,000 count concerns the Matenadaran collection of Armenian manuscripts. That number refers to manuscripts in a library, not to patristic Bible commentaries as a genre. The most defensible conclusion is therefore this: 17,000 is not a reliable single-number description of the patristic and medieval exegetical corpus unless one first specifies the unit being counted. If the unit is medieval Latin commentary records, the true figure may be closer to 24,000 in the Trier repertory. If the unit is patristic works overall, the figure changes radically. If the unit is manuscripts, it can rise or fall by tradition and work-family. If the unit is verse-level references, it is vastly larger still. 61 62 3 63 • 11 • 50 • 64 • 65 • 5 • 66 67 12