Monthly Archives: January 2025

Who Killed Marrucinus Asinius?

Posted on by

My fourth semester Latin students have just started reading Catullus, and I wanted a sight-reading activity that would reinforce the vocabulary from poems 12 (Marrucine Asini) and 13 (Cenabis bene). I asked ChatGPT to create a brief murder mystery in Latin involving the characters from the poems. Here is the yarn it spun. Prompt and Latin critique to follow . . .

Quis Marrucinum Asinium Necavit?

Marrucinus Asinius, notus linteorum fur, subito mortuus repertus est in villa Catulli. Corpus eius iacebat in tricliniis, inter vinum et reliquias cenae. Fabullus et Veranius, amici Catulli, stupefacti spectabant. Candida puella, quae nuper ad convivium venerat, fletu oculos suos tegebat. Catullus ipse, pallidus et tacitus, stabat prope corpus.

Fabullus dixit: “O di immortales! Quis hoc fecit? Marrucinus mihi linteum subripuit, sed num homicidium meruit?”

Veranius respondit: “Estne hoc opus irati Catulli? Tu saepe minatus es, Catulle, hendecasyllabos trecentos componere contra eum. Fortasse ira tua excessit in caedem.”

Catullus oculos sublatis vocavit: “Non ego! Quamvis illum furti accusaverim, nunquam tamen violentiâ usus sum.”

Tum Fabullus: “Sed quid de candida puella?”

Puella, lacrimis plena, dixit: “Marrucinus mihi insultavit in convivio! Sed num credetis me, fragilem feminam, eum necavisse? Non ego id feci!”

Omnes oculos in Fabullum verterunt. “Fabulle, quid dicis?” inquit Veranius. “Tu saepe Marrucinum avaritiae accusasti.”

Fabullus caput movere coepit: “Certe fur erat! Sed nec ego necavi! Certe linteum meum mihi carum erat, sed hoc non est causa caedis.”

Tum Catullus ad corpus accessit et rem miram animadvertit. “Ecce! Linteum Saetabum prope corpus iacet. Hoc linteum mihi a vobis donatum est, Fabulle et Verani! Num hoc est signum?”

Puella subridens respondit: “Signum, fortasse. Sed cur linteum afferre me putatis? Potius, spectate versum in pariete scriptum: ‘Fur et iniustus iacet. Hendecasyllabi ultorem iustitiam tulerunt.’”

Omnes stupuerunt. Veranius tremens dixit: “Hendecasyllabi? Num versus tuae irae, Catulle, animati facti sunt?”

Catullus suspiravit: “Fortasse versus mei plus potentiae habent quam ego imaginatus sum.”

Fabullus: “Itaque, non puella, non ego, non tu, sed hendecasyllabi ipse Marrucinum necaverunt?”

Puella ridens: “Quis scit? Forsitan poesis ipsa vindicem aequitatemque tulerit.”

Omnes tacuerunt, linteum et corpus spectantes. Vino mox impleverunt pocula et memoriam Marrucini – furis invisi sed hominis tamen – saltem modico risu honoraverunt.

The prompt was as follows. I was careful to specify my role, what I was trying to do, what I wanted the AI to do, and the format in which I wanted the results.

I am a Latin teacher. I am trying to create an easy Latin story for my students, who are at the intermediate level. I want the story to be a murder mystery, entitled, Quis Marrucinum Asinium necavit, or “Who Killed Marrucinus Asinius?” It should be in Latin, and the characters should include the people mentioned in Catullus’ poems 12 and 13. I will put those poems in the prompt. One of the characters, the victim, is Marrucinus Asinius, a napkin thief. I want the story to be a murder mystery in which Marrucinus shows up dead and it is unclear who has killed him. It could be the beautiful girl (candida puella) mentioned in poem 13, or it could be Catullus’ friends, Varnius or Fabullus, or it could be Catullus himself, or it could be the hendecasyllables, or lines of invective poetry mentioned in poem 12. I would like to use as much of the vocabulary that appears in poems 12 and 13 as possible, but it is also fine to use other common Latin words. The story should be about 300-500 words long.

Here is poem 12:

Marrucine Asini, manu sinistra

non belle uteris in ioco atque vino:       

tollis lintea neglegentiorum.    

hoc salsum esse putas? fugit te, inepte:           

quamvis sordida res et invenustast.    

non credis mihi? crede Pollioni             

fratri, qui tua furta vel talento  

mutari velit: est enim leporum

differtus puer ac facetiarum.   

quare aut hendecasyllabos trecentos 

exspecta, aut mihi linteum remitte;     

quod me non movet aestimatione,       

verumst mnemosynum mei sodalis.    

nam sudaria Saetaba ex Hiberis           

miserunt mihi muneri Fabullus             

et Veranius: haec amem necessest     

ut Veraniolum meum et Fabullum.

Here is poem 13:

 Cenabis bene, mi Fabulle, apud me   

paucis, si tibi di favent, diebus,             

si tecum attuleris bonam atque magnam        

cenam, non sine candida puella          

et vino et sale et omnibus cachinnis.

haec si, inquam, attuleris, venuste noster,      

cenabis bene: nam tui Catulli 

plenus sacculus est aranearum.          

sed contra accipies meros amores      

seu quid suavius elegantiusvest:

nam unguentum dabo, quod meae puellae     

donarunt Veneres Cupidinesque;         

quod tu cum olfacies, deos rogabis,    

totum ut te faciant, Fabulle, nasum.

What the AI produced is not very elegant Latin. It knows to put the verb at the end of the sentence by default. But there are traces of English word order and style translated directly into Latin. ‘Plus … quam ego imaginatus sum’ sounds like English to me. I had to ask it what it meant by ‘Hendecasyllabi ultorem iustitiam tulerunt.’ It obliged by explaining:

“The hendecasyllables have brought avenging justice.”

It implies that the poetic lines themselves (the hendecasyllabic verses) acted as instruments of justice or vengeance against Marrucinus. It adds a mysterious, almost supernatural twist to the story!

But iustitiam ferre is not a thing, per TLL. I think the AI means something like supplicium sumpsit.

On the other hand, maybe I’m missing something, but I see only one genuine mistake: oculos (bolded above) should be oculis. Credetis (also bolded) should probably be creditis, but it’s not an outright error.

I was planning to use this to have the students spot and correct errors, but I really can’t find any other outright mistakes, as opposed to instances of sub-par Latin. Wow.

 

Digital Commentaries and Generative AI

Posted on by

A fascinating recent paper by Sarah Abowitz, Alison Babeu, and Gregory Crane, all of Tufts University, asks whether “general advances in machine learning could power alternative digital aids” to reading foreign-language sources more easily, “requiring less labor” than the immense effort required to create traditional print commentaries. In other words, how can we best use technology to transcend the limits of print in helping readers of classical works in the digital age? It’s a question at the heart of DCC, and of the other digital commentary projects they mention, the Ajax Multi-commentary Project and New Alexandria. After surveying classical commentary traditions past and present, print and digital (with some very kind words about DCC) they discuss a study conducted on sample commentaries aimed at different audiences on sections from Thucydides’ History and from the Iliad—Book 6 of both works. The authors do not consider the rich tradition of Hebrew and Christian exegesis, but their comments on the classical tradition could perhaps be extended to other traditions as well.

Unknown artist/maker, illuminator, Elijah ben Meshallum, scribe, Elijah ben Jehiel, scribe, et al. Decorated Text Page from the Rothchild Pentateuch, 1296. J. Paul Getty Museum, Los Angeles

For the study, two coders tagged the comments and categorized them by function into six varieties: syntactic aid, translation, semantic aid, inconsistency alerts (which informed the reader of changes the editor made to the original text), stylistic claims (which assert that a certain linguistic feature occurs in a certain way in the text, for example, “this is one of Nicias’ favorite adjectives”), and finally reference pointers, which refer to any specific reference work, primary source, or scholarly paper.

The authors acknowledge that these categories are not watertight, and that quite a few comments straddled more than one of the categories. Unsurprisingly, commentaries that are oriented towards scholars have a higher percentage of reference pointers and less the way of translation and syntactic aids. Those aimed at students have more in the way of syntactic aids and translation. The high number of bare reference pointers that are characteristic of scholarly commentaries (I call them “cf.” notes and attempt to rigorously exclude them from DCC) are often dead ends in an open digital medium, since the material referred to is copyrighted and behind paywalls.

The authors point out that there is very little research about user experience of commentaries, of the kind that would be standard procedure in creating any self-respecting digital interface. What do we really know about how students use commentaries on average, and which types of notes are most helpful? We all have opinions on this topic, but apparently little or nothing has been done formally to investigate the question.

DCC’s practice is now to road-test commentaries with students before publication. In some cases, students are involved in choosing notes from existing public domain commentaries for variorum editions. A new cohort of DCC high school interns is on the way for summer 2025. We’re not gathering any data about student preferences from this process, but we could.

Another interesting point the authors make, coming from a computer science perspective, is that the stylistic claims in commentaries are often based on data (for example the number of occurrences of a given word or phrase), yet classical commenters almost never provide the data to back these assertions up (Some notes in the Cambridge Green & Yellow series do, and Ronald Syme’s Tacitus is a notable pre-digital exception). In a digital medium, it would be quite possible to provide the data, for example using treebank data.

The goals of this kind of discussion, it seems to me, are two: first to serve readers better, and second to use computational means to create more content with less labor. To what extent can artificial intelligence and generative AI aid in this enterprise? This paper suggested to me an interesting approach to testing and moving forward. The creating of a typology of comments is an important advance. This could be refined by analyzing a larger sample of older commentaries, akin to the corpus collected for the Ajax project on Sophocles’ Ajax, though these are all very scholarly, and not very useful for most readers, in my opinion. It would be better to work with school commentaries where multiple parallel school commentaries exist, say Caesar, Cicero, or Vergil, Xenophon, Homer, or Lysias.

Once a refined and well-understood typology of notes is ready, the types can be evaluated in terms of their relative utility for different audiences, and serve as a basis for creating prompts for generative AI. A specific, well-designed prompt might elicit from AI a certain type of comment, and a combination of those could be used to create a draft commentary on new texts, to be evaluated and edited by humans.  

Another key piece of the puzzle is something that Gregory Crane mentioned at the recent SCS panel, the fact that Hathi trust now has marvelously good multilingual OCR in its back end. In theory it will be possible to do a much larger harvesting of existing public domain commentaries, tag and use that data to create a more extensive collection of note types and do a kind of sifting operation in which the users select the notes that they find particularly helpful. The knowledge thus gained could be used on creative AI prompts. Obviously, anything produced by generative AI would have to be extensively edited by humans. But getting us part way would be extremely helpful. My own experiments on generating vocabulary lists with ChatGPT seem promising.

SCS 2025 Panel on Using AI in Classical Philology

Posted on by

Every year the Digital Classics Association (DCA) sponsors a panel at the meetings of the Society for Classical Studies (SCS), and this year, appropriately, it was all about the potential of LLMs and AI in classical philology. Organized by Neil Coffee, the driving force behind DCA, it was a star-studded panel. Here is (was) the program:

Opening Up Classics with AI (organized by the Digital Classics Association)
Neil Coffee, University at Buffalo, SUNY, Organizer

  1. Neil Coffee, University at Buffalo, SUNY
    Introduction
  2. Samuel Huskey, University of Oklahoma
    Opening Up Bottlenecks in Digital Classics Workflows with Human-in-the-Loop AI
  3. Patrick Burns, New York University
    Prompt Engineering for Latin Teachers
  4. Edward Ross, University of Reading, and Jackie Baines, University of Reading
    Generative Image AI and Teaching Classics: A Case of Exaggeration
  5. Gregory Crane, Tufts University
    AI, Machine Actionable Publication and Assigning Credit
  6. Joseph Dexter, Harvard University, and Pramit Chaudhuri, University of Texas at Austin
    Benchmarking Generative AI Models for Classical Literary Criticism
Slide listing Gregory Crane's goals for his talk at the SCS

Gregory Crane, Tufts University
AI, Machine Actionable Publication and Assigning Credit

The abstracts of the talks are posted here, so I won’t try summarize them. My favorite quote came from the ever-polemical Gregory Crane, who referred to the monographs being sold nearby in the SCS book publishers’ display as “a dark archive,” and said “publications/datasets matter insofar as they fire the human mind.” That is a scholarly goal I can get behind, firing the human mind.

Huskey is working on the gathering comprehensive metadata for the Digital Latin Library, sucking in library records from all over the world and trying to disambiguate author names and work titles, many of which have multiple variants, confusing overlaps, and vagueness in the existing records (opera omnia? opera selecta? Bucolica? Eclogae?)

Burns is working on trying to create extensive reading material for Latin learners, as we go from the extreme scarcity of comprehensible texts for beginners to a world where we can have essentially infinite amounts of Latin pitched at any level. Charmingly, he had an LLM create a story about Odysseus and the Cyclops from the perspective of the sheep. From this talk I learned that prompts can be very large. A human will be confused by a question that is 300 pages long; AIs can easily take it in and synthesize. His main message is you can get a basic understanding of how this things work without being a computer scientist, and it is helpful to have such an understanding. 

Ross and Baines are keeping track of AI-generated images that have something to do with the ancient world, and ferreting out distorted history, incorrect information, and modern biases. They showed an amusing image of “Nike the Greek Goddess” flying around wearing a pair of Nike sneakers. Images of Hades draw extensively from Disney’s Hercules. They believe scholars have a duty to keep track of the craziness that is out there, if only to help the image tools get gradually more historically accurate.

Dexter and Chaudhri just finished teaching a seminar on Latin literary history using only fragmentary authors, and are trying to use AI to craft a new narrative about Latin literary history based on this material.

Crane wants to leverage AI to give people without extensive knowledge of historical languages better access to the classics of the world, the whole world, through enhanced translations and reference tools, and to serve audiences in their own languages (e.g. Persian), not just English. I was particularly taken with his effort to us machine translation to translate the examples in Kuhner-Gerth’s Ausführliche Grammatik der griechischen Sprache into English, thus unlocking this fundamental reference work for a broader audience. He’s training AI to figure out which of the 26 uses of ὡς is active in a particular passage of Greek. He also pointed out the the multilingual OCR at Hathi Trust is off-the-chain good at this point.

What struck me was the way that using AI tools requires scholars to be explicit about their goals, what they really want to do, in a way that writing a journal article does not.  These papers used AI tools for different, all legitimate, philological and scholarly goals. Do you want to

  • critique historical bias and inaccuracy on the web? (Ross & Baines)
  • tell a story about literary history? (Dexter and Chaudhuri)
  • help people learn Latin? (Burns)
  • catalogue published texts? (Huskey)
  • attribute passages correctly? (Dexter and Chaudhuri)
  • fire the human mind? (Crane)

AI can help. Notably, none of these goals is rewarded by the academic world as currently constituted. Which is one more reason I respect these scholars for doing interesting work despite the professional incentives to churn out another article or book for the dark archive.

I went into this panel rather repelled by AI, more aware of it as a tool for cheating on college writing assignments, and a potential menace to humanity, than as a potential aid in my beloved philology. I came out intrigued with the possibilities and wanting to try to apply it to the workflows of DCC (see this post on my first attempts to create DCC style vocabulary lists with ChatGPT and Claude).

Yuval Noah Harari’s fascinating 2024 book Nexus convinced me that there is no pre-determined end to the AI story, and that we need to be actively engaged in thinking about it and guiding its trajectory for human goals. Harari, a historian, talks about the ways that every new information technology brought good things and bad things. Printing enabled both the scientific revolution and witch hunts. What matters is how we use it and shape it. These papers all showed scholarly uses of AI that seem to me both interesting and productive.

AI on the Bravery of Women

Posted on by

Fresh off the stimulating Digital Classics Association panel at the 2025 SCS on uses of AI in classical scholarship, I decided to give ChatGPT 4 and Claude 3.5 Sonnet a crack at one of the central tasks of DCC: creating accurate vocabulary lists. 

Consider a work in our pipeline, Plutarch’s On the Bravery of Women, which begins as follows:

περὶ ἀρετῆς, ὦ Κλέα, γυναικῶν οὐ τὴν αὐτὴν τῷ Θουκυδίδῃ γνώμην ἔχομεν. ὁ μὲν γάρ, ἧς ἂν ἐλάχιστος ᾖ παρὰ τοῖς ἐκτὸς ψόγου πέρι ἢ ἐπαίνου λόγος, ἀρίστην ἀποφαίνεται, καθάπερ τὸ σῶμα καὶ τοὔνομα τῆς ἀγαθῆς γυναικὸς οἰόμενος δεῖν κατάκλειστον εἶναι καὶ ἀνέξοδον. ἡμῖν δὲ κομψότερος μὲν ὁ Γοργίας φαίνεται, κελεύων μὴ τὸ εἶδος ἀλλὰ τὴν δόξαν εἶναι πολλοῖς γνώριμον τῆς γυναικός: ἄριστα δ᾽ ὁ Ῥωμαίων δοκεῖ νόμος ἔχειν, ὥσπερ ἀνδράσι καὶ γυναιξὶ δημοσίᾳ μετὰ τὴν τελευτὴν τοὺς προσήκοντας ἀποδιδοὺς ἐπαίνους. διὸ καὶ Λεοντίδος τῆς ἀρίστης ἀποθανούσης, εὐθύς τε μετὰ σοῦ τότε πολὺν λόγον εἴχομεν οὐκ ἀμοιροῦντα παραμυθίας φιλοσόφου, καὶ νῦν, ὡς ἐβουλήθης, τὰ ὑπόλοιπα τῶν λεγομένων εἰς ‘τὸ μίαν εἶναι καὶ τὴν αὐτὴν ἀνδρὸς καὶ γυναικὸς ἀρετὴν προσανέγραψά σοι, τὸ ἱστορικὸν ἀποδεικτικὸν ἔχοντα καὶ πρὸς ἡδονὴν μὲν ἀκοῆς οὐ συντεταγμένα εἰ δὲ τῷ πείθοντι καὶ τὸ τέρπον ἔνεστι φύσει τοῦ παραδείγματος, οὐ φεύγει χάριν ἀποδείξεως συνεργὸν ὁ λόγος οὐδ᾽ αἰσχύνεται
ταῖς Μούσαις
τὰς Χάριτας συγκαταμιγνὺς
καλλίσταν συζυγίαν,
ὡς Εὐριπίδης φησίν, ἐκ τοῦ φιλοκάλου μάλιστα τῆς ψυχῆς ἀναδούμενος τὴν πίστιν.
 
In the Loeb translation of Frank Cole Babbit

Regarding the virtues of women, Clea, I do not hold the same opinion as Thucydides. For he declares that the best woman is she about whom there is the least talk among persons outside regarding either censure or commendation, feeling that the name of the good woman, like her person, ought to be shut up indoors and never go out. But to my mind Gorgias appears to display better taste in advising that not the form but the fame of a woman should be known to many. Best of all seems the Roman custom, which publicly renders to women, as to men, a fitting commemoration after the end of their life. So when Leontis, that most excellent woman, died, I forthwith had then a long conversation with you, which was not without some share of consolation drawn from philosophy, and now, as you desired, I have also written out for you the remainder of what I would have said on the topic that man’s virtues and woman’s virtues are one and the same. This includes a good deal of historical exposition, and it is not composed to give pleasure in its perusal. Yet, if in a convincing argument delectation is to be found also by reason of the very nature of the illustration, then the discussion is not devoid of an agreeableness which helps in the exposition, nor does it hesitate
To join
The Graces with the Muses, A consorting most fair,
as Euripides says, and to pin its faith mostly to the love of beauty inherent to the soul.

I gave both AIs a lengthy prompt, similar to what I would say to a human if I were tasking her with creating a vocabulary list:

I am going to attach a .txt file with some Ancient Greek. I want you to create a vocabulary list for the text. Each entry in the list should contain the standard dictionary form of the token, followed by an English definition appropriate to the context. For example, given the token γυναικῶν, the entry in the list should read “γυνή γυναικός, ἡ: woman.” γυνή γυναικός, ἡ is the dictionary form. This token is a noun, so it includes the nominative singular, γυνή, followed by the genitive singular, γυναικός, followed by a comma, then the feminine form of the definite article (ἡ), indicating it is a feminine noun. After that comes a colon, which separates the dictionary form from the English definition. I used “woman” in this example rather than “wife,” which is another possible definition. In the context from which I took the example, the author is discussing women in general, not just wives. Here is a second example, this time for a verb. The token is ἔχομεν. The dictionary form is “ἔχω, ἕξω or σχήσω, 2 aor. ἔσχον, ἔσχηκα, impf. εἶχον” . I derived this dictionary form from the Dickinson College Commentaries site https://dcc.dickinson.edu/greek-core-list . Use that if you can, but that doesn’t have all words. For words not in that source, use the fuller list in the Grieks Nederlands dictionary available on the site Logeion https://logeion.uchicago.edu/ . If the dictionary form there is very long, try to simplify it based on the format of the Dickinson College Commentaries list and the following list of further examples, which includes some adjectives, adverbs, and verbs.
ὑπόληψις -εως, ἡ: opinion, assumption

ὁρμή -ῆς, ἡ: impulse

ὄρεξις -εως, ἡ: desire

ἔκκλισις -εως, ἡ: aversion, avoidance

κτῆσις -εως, ἡ: possession, property, property

δόξα -ης, ἡ: opinion; reputation

1.2

ἀκώλυτος -ον: unhindered

ἀπαραπόδιστος -ον: unimpeded, unobstructed

ἀσθενής -ές: weak, powerless

δοῦλος -η -ον: slavish, servile

κωλυτός -ή -όν: hindered

ἀλλότριος -α -ον: not one’s own, under the control of others

1.3

ἐμποδίζω -ποδιῶ -επόδισε: to hinder, frustrate

πενθέω -ήσω ἐπένθησα: to mourn, to suffer pain

ταράσσω ταράξω ἐτάραξα: to trouble, disturb

μέμφομαι μέμψομαι ἐμέμφθην: to blame

ἀναγκάζω -άσω: to force, constrain

οὐδέποτε: not ever, never

ἐγκαλέω ἐγκαλῶ ἐνεκάλεσα: to rebuke, reproach + dat.

ἄκων -ουσα -ον: involuntarily, against one’s will

βλάπτω βλάψω ἔβλαψα: to harm, hinder, disable

βλαβερός -ά -όν: harmful

1.4

τηλικοῦτος -αύτη -οῦτον: such great

ἐφίημι ἐφήσω ἐφῆκα: to aim at, to long for, desire + gen.

μετρίως (adv.): moderately

ἅπτω ἅψω ἥψα: to grasp, obtain, engage in, undertake (+ gen)

παντελῶς (adv.): absolutely, completely

ὑπερτίθημι ὑπερθήσω ὑπερέθηκα: to defer, put off

πλουτέω -ήσω ἐπλούτησα: to be rich or wealthy

ἀποτυγχάνω ἀποτεύξομαι ἀπέτυχον: to fail in attaining, miss

ἐλευθερία -ας, ἡ: freedom

εὐδαιμονία -ας, ἡ: prosperity; true, full happiness

περιγίγνομαι περιγενήσομαι περεγενόμην: to be a result or consequence, to arise

You will notice that adverbs sometimes have “(adv.)” at the end of the dictionary form. That is desirable. Note also that when more than one English definition is given, sometimes those English definitions are separated by commas, and sometimes by semi-colons. The commas separate English definitions that are near-synonyms. The near synonyms are grouped together and separated from substantially different possible English definitions with a semi-colon. Notice also that there are milestones separating groups of entries that derive from a single paragraph or sentence, such as “1.1,” “1.2.” If the text I input seems to have section headers, either numerical like that, or a title, please include those on a separate line above the list of words for that section. Section headers often include English words.

That was the prompt. Claude made a nice summary of my prompt:

I understand you want me to create a vocabulary list from Ancient Greek text, following specific formatting rules. Each entry should have:

1. Dictionary form (using Dickinson College Commentaries when possible, otherwise simplified entries from Logeion)
2. English definition(s) appropriate to context
3. Special formatting for different parts of speech (e.g., adding “(adv.)” for adverbs)
4. Grouping similar English definitions with commas, different meanings with semicolons
5. Including section headers when present
6. For nouns: nominative + genitive + article
7. For adjectives: all gender endings
8. For verbs: principal parts in simplified format

I see you’ve provided helpful examples to guide the formatting. However, I don’t see the actual Greek text file you mentioned wanting to attach. Could you please share the text you’d like me to analyze? Once you do, I’ll create a vocabulary list following these specifications.

I will paste the raw results below. I also have the results in a Word file, edited with track changes to show all the improvements I would make. If anybody wants that I can supply it. Here is my summary analysis.

  1. Both AIs did a superb job of crafting dictionary forms to my specifications.
  2. There were occasional errors of lemmatization, e.g. ἐπαινέω for ἔπαινος, σύζυγος for συζυγία, φιλοκαλέω for φιλόκαλος, ἄμοιρος for ἀμοιρέω. Claude was somewhat better at lemmatization than GPT, but both were very good. GPT astutely picked up an instance of ὁ μὲν … ὁ δέ: “on the one hand … on the other hand (used to contrast ideas).”
  3. Both helpfully omitted very common words, though I had given no instructions in that regard.
  4. Both omitted a number of uncommon words, 9 in the case if GPT, 15 in the case of Claude. Claude missed all the proper names.
  5. The definitions were generally very good. Rare meanings like παράδειγμα = “illustration” and ἀπόδειξις = “exposition” were missed by both AIs
  6. GPT tried to give a case-specific definition for μετά twice, and was wrong both times. On μετὰ τὴν τελευτὴν it offered μετά + gen.: “with, among”; on μετὰ σοῦ it offered  μετά + acc.: “after.”

Very impressive, I would say, despite the errors. Just in terms of the formatting of dictionary lemmas, this will be a huge time saver. I don’t think we will abandon our usual workflow with the Bridge, but this is very much a viable option going forward.

The main thing I learned at the DCA panel is that you can put in huge prompts with elaborate instructions, no problem. Boy, do I feel slow now when I edit vocab lists, like a snail trying to keep up with a Lamborghini.

Ok, here are the raw results from GPT:

ἀρετή -ῆς, ἡ: virtue, excellence

γυνή γυναικός, ἡ: woman

Θουκυδίδης -ου, ὁ: Thucydides (name)

γνώμη -ης, ἡ: opinion, judgment, intelligence

ἔχω, ἕξω or σχήσω, 2 aor. ἔσχον, ἔσχηκα, impf. εἶχον: to have, hold, possess; to be able to

ὁ μὲν … ὁ δέ: on the one hand … on the other hand (used to contrast ideas)

ὀλίγος -η -ον: small, little, few

ψόγος -ου, ὁ: blame, criticism

ἔπαινος -ου, ὁ: praise, commendation

καθάπερ: just as, as if

σῶμα σώματος, τό: body

ὄνομα ὀνόματος, τό: name

ἀγαθός -ή -όν: good, noble, brave

οἴομαι, οἰήσομαι, ᾠήθην: to think, suppose, imagine

δεῖ: it is necessary, must

κατάκλειστος -ον: closed, shut in

ἀνέξοδος -ον: inaccessible, closed off

κομψός -ή -όν: refined, elegant, clever

Γοργίας -ου, ὁ: Gorgias (name)

κελεύω, κελεύσω, ἐκέλευσα: to order, command

εἶδος -ους, τό: form, shape, appearance

δόξα -ης, ἡ: opinion, glory, reputation

δημοσίᾳ (adv.): publicly, in public

μετά + gen.: with, among

τελευτή -ῆς, ἡ: end, death

ἀποδίδωμι, ἀποδώσω, ἀπέδωκα: to give back, return, pay

ἐπαινέω, ἐπαινέσω, ἐπῄνεσα: to praise, commend

Λεοντίς -ίδος, ἡ: Leontis (name)

ἀποθνῄσκω, ἀποθανοῦμαι, ἀπέθανον: to die

εὐθύς: immediately, at once

μετά + acc.: after

πολύς, πολλή, πολύ: much, many

λόγος -ου, ὁ: word, speech, discourse

παραμυθία -ας, ἡ: consolation, comfort

φιλόσοφος -ου, ὁ/ἡ: philosopher

βουλεύω, βουλεύσω, ἐβούλευσα: to deliberate, plan

ὑπόλοιπος -ον: remaining, left over

λέγω, λέξω or ἐρῶ, εἶπον: to say, speak, tell

μία -ας, ἡ: one, single

ἀνήρ ἀνδρός, ὁ: man, husband

καταγράφω, καταγράψω, κατέγραψα: to record, write down

ἱστορικός -ή -όν: historical

ἀποδεικτικός -ή -όν: demonstrative, explanatory

ἡδονή -ῆς, ἡ: pleasure, delight

ἀκοή -ῆς, ἡ: hearing, sense of hearing

συντάσσω, συντάξω, συνέταξα: to arrange, compose

φύσις -εως, ἡ: nature, character

παράδειγμα -ατος, τό: example, pattern

λόγος -ου, ὁ: reason, word, account

χαρίς -ίτος, ἡ: grace, favor, kindness

Μοῦσα -ης, ἡ: Muse (goddess of art and inspiration)

Χάρις -ιτος, ἡ: Grace (one of the Graces)

καλλίστος -η -ον: most beautiful

σύζυγος -ον: yoked together, paired

φιλοκαλέω: to love beauty, to be elegant

ἀναδέω, ἀναδήσω, ἀνέδησα: to bind up, to crown

And here are the raw results from Claude:

Let me create a vocabulary list for this text from Plutarch’s “On the Virtues of Women.”

Title: On the Virtues of Women (Introduction)

ἀρετή -ῆς, ἡ: excellence, virtue
γυνή γυναικός, ἡ: woman
γνώμη -ης, ἡ: opinion, judgment
ἔχω, ἕξω or σχήσω, 2 aor. ἔσχον: to have, hold
ἐλάχιστος -η -ον: least, smallest
ψόγος -ου, ὁ: blame, criticism
ἔπαινος -ου, ὁ: praise
λόγος -ου, ὁ: word, speech; account
ἄριστος -η -ον: best, excellent
ἀποφαίνω -φανῶ -έφηνα: to show, declare, present
σῶμα -ατος, τό: body
ὄνομα -ατος, τό: name
ἀγαθός -ή -όν: good, noble
κατάκλειστος -ον: shut up, confined
ἀνέξοδος -ον: not going out, staying at home
κομψός -ή -όν: refined, clever
φαίνω φανῶ ἔφηνα: to show, appear
κελεύω -σω ἐκέλευσα: to order, command
εἶδος -εος, τό: form, appearance
δόξα -ης, ἡ: reputation; opinion
γνώριμος -ον: well-known, familiar
Ῥωμαῖος -α -ον: Roman
νόμος -ου, ὁ: law, custom
δημόσιος -α -ον: public
τελευτή -ῆς, ἡ: end, death
προσήκω -ξω -ῆξα: to be fitting, proper
ἀποδίδωμι -δώσω -έδωκα: to give back, render
ἄμοιρος -ον: without share in (+ gen.)
παραμυθία -ας, ἡ: consolation, comfort
φιλόσοφος -ον: philosophical
ὑπόλοιπος -ον: remaining, rest
ἀποδεικτικός -ή -όν: demonstrative, proving
ἡδονή -ῆς, ἡ: pleasure
ἀκοή -ῆς, ἡ: hearing
συντάσσω -τάξω -έταξα: to arrange, compose
πείθω πείσω ἔπεισα: to persuade
τέρπω τέρψω ἔτερψα: to delight, please
παράδειγμα -ατος, τό: example, model
φεύγω φεύξομαι ἔφυγον: to flee, avoid
χάρις -ιτος, ἡ: grace, charm; gratitude
συνεργός -όν: helping, cooperating
αἰσχύνω -νῶ ᾔσχυνα: to shame
Μοῦσα -ης, ἡ: Muse
Χάρις -ιτος, ἡ: Grace (one of the Graces)
συγκαταμίγνυμι -μίξω -έμιξα: to mix together, blend
καλός -ή -όν: beautiful, noble
συζυγία -ας, ἡ: union, pairing
φιλόκαλος -ον: loving beauty
ψυχή -ῆς, ἡ: soul
ἀναδέω -δήσω -έδησα: to bind up, crown
πίστις -εως, ἡ: trust, faith; proof