The Pliny vs. Caesar Vocabulary Showdown (A Pliny Appreciation Guide, Part 5)

When I asked the AP Latin teachers’ Facebook group about their misgivings with Pliny, vocabulary was a common theme:

“While I know most of the vocab will be glossed, it was distracting that we couldn’t read it without having to look up most of the vocab. Caesar’s vocab and grammar were much easier for my students.”

“I don't know word counts off hand, but I know Steadman’s initial response was that the move would make things harder for students because there's far less high frequency vocab. That seems correct to me, though I don't know the exact numbers. I'm sure students will have more trouble with Pliny vocab than they ever did with Caesar vocab.” (I was unable to find the referenced post on Geoffrey Steadman’s blog.)

Well, I’ve got the numbers, so let’s see what we find!

The first thing I realized is that comparing the old Caesar AP readings to the new Pliny AP readings doesn’t work because of the discrepancy in overall word count. To make a fair comparison, it became necessary to take a sample of Caesar that was about the same length as the AP Pliny passages (~2600 words). I also took equivalent samples from eight other prose authors: Cicero, Sallust, Nepos, Livy, Petronius, Seneca, Tacitus, and Suetonius.^[1] Then, I ran all these passages through The Bridge and played around with the data a bit, looking for major discrepancies between Caesar and Pliny.

To my dismay, I did actually find two important areas in which Caesar clearly outdoes Pliny. The first is in overall lexical density: the number of unique words in the passage divided by the total number of words in the passage. Caesar wins the prize for the least lexically dense of all ten authors. Pliny, on the other hand, comes in 7th, right ahead of his buddies Tacitus and Suetonius.

Author	Total words	Unique words	Vocab density
Caesar	2650	799	0.302
Cicero	2590	804	0.310
Nepos	2641	882	0.334
Seneca	2612	910	0.348
Sallust	2597	915	0.352
Livy	2473	920	0.372
Pliny	2658	1085	0.408
Petronius	2523	1069	0.424
Tacitus	2667	1157	0.434
Suetonius	2630	1272	0.484

The other metric is a related concept. Within each selection, there are words the author uses only once in that selection (hapax legomena). It turns out that, if we look only at the AP syllabus, Pliny uses quite a lot of words only once and never returns to them again. In fact, 1 in 4 of the total words on the AP Pliny syllabus is a hapax legomenon. Contrast that with Caesar, who comes in at 1 in 6—again, the lowest out of the authors in the sample.

Author	Total words	Hapax legomena	% hapax
Caesar	2650	422	15.9%
Cicero	2590	443	17.1%
Nepos	2641	505	19.1%
Sallust	2612	516	19.9%
Seneca	2597	564	21.6%
Livy	2473	534	21.6%
Pliny	2658	678	25.5%
Tacitus	2523	715	26.8%
Petronius	2630	707	28.0%
Suetonius	2667	880	33.5%

So it seems that the reason that some teachers and students were getting frustrated at all the words they had to look up is because Pliny recycles words much less often than Caesar. Contrary to assumptions, it’s actually not really a matter of high-frequency vs. low-frequency words. If we count the percentage of total words that are among the 100 most commonly used (based on The Bridge’s corpus frequency rankings), Caesar and Pliny are both in the middle of the pack (4th and 6th out of 10, respectively). For both authors, about two-thirds of their words are in the top 600 most frequent, and about three-quarters in the top 1,000. So it’s not really about the kinds of words Caesar uses, but rather how often he repeats them.

Author	% of total words in top 100	% of total words in top 300	% of total words in top 600	% of total words in top 1000
Cicero	49.5%	63.8%	74.1%	80.0%
Seneca	48.2%	62.9%	72.5%	78.8%
Nepos	43.7%	58.2%	68.9%	76.3%
Caesar	43.4%	58.2%	68.7%	76.5%
Sallust	38.9%	56.6%	67.4%	75.2%
Pliny	38.8%	54.2%	64.1%	73.0%
Livy	37.2%	50.8%	62.5%	68.6%
Petronius	36.9%	48.5%	56.5%	62.9%
Suetonius	33.2%	45.7%	55.3%	63.6%
Tacitus	32.7%	46.9%	57.6%	67.1%

As a final thought, observe that Pliny’s lexicon is really all that peculiar when you compare it to the wide world of Latin prose. In fact, in all of the above metrics, Pliny is closer to the median than Caesar is. Lexically speaking, Caesar is the aberration in Roman literature, not the norm.

So, what do we do with this?

Now, at this point I could go on and on about how Pliny varies his vocabulary for literary effect and how the brevity of the epistolary genre demands a certain lexical variety. I could point out that it’s silly to hold Pliny to the lexical standards of Caesar, an author not only writing in a completely different genre but chronologically as far removed from Pliny as we are from Dickens. I could even say that “a blend of such ‘modern’ characteristics combined with a firm basis of classical lexicon to produce one of the ingredients of that ‘refined colloquialism’ peculiar to the letters.”^[2]

But I’m not going to do that. That’s not what the anti-Plinians want to hear. Their objection is not that Pliny’s letters are no good (well, sometimes it is). Their objection is that their letters are no good for learners of Latin. And as someone who did his master’s thesis on sheltered-vocabulary Latin novellas, I agree that the more a text recycles words, the better the text is as a teaching tool. In that respect, Caesar is a better teaching tool than Pliny.

But I come to bury Caesar, not to praise him^[3]—I believe that, in practically all other respects, his value as a text for Latin learners is grossly overstated.

You see, there’s a tendency I’ve observed in AP Latin folk—both its teachers and the curriculum itself—to view Latin texts as valuable only insofar as they represent exemplars of the Latin language. I’ve seen this firsthand in the AP Latin teachers’ Facebook group. While discussing the merits of Pliny vs. Caesar, one teacher gave an example of a sentence of Caesar, which, more than any sentence of Pliny, “rewards the effort put in to translate it.” When I looked up the citation (DBG 5.27, about a fairly banal diplomatic meeting), I asked this teacher what was so remarkable about it, thinking I was missing some key context. The teacher responded, “Caesar gives an 11-sentence speech in oratio obliqua which, though difficult, manages never to stray from normal rules of syntax.” The content of the sentence was never addressed, only its linguistic qualities. I found this awfully telling: the payoff of reading this sentence is the language and grammar of the sentence, while the meaning—the entire reason Caesar wrote the sentence in the first place—is relegated to little or no importance. (Still, credit where it’s due: this teacher kept an open mind and revisited some of their first impressions of Pliny, so I do appreciate that!)

But the general chauvinism for Caesar, I think, stems from this idea that the most important thing about a Latin text is its linguistic features. To AP die-hards, Caesar is the perfect Latin author: he uses the same words all the time, he never departs from textbook grammatical usage, he writes nice, neat periodic sentences where everything fits together just so. In many ways, this represents the core of the AP Latin philosophy: it is not the author’s message that matters, it’s the Latin language itself.

If we are judging Roman literature by the sole question of which one is the absolute easiest for high school students on a linguistic level, then sure, let’s all read Caesar. Let’s ignore the fact that the genre of military commentary is engaging only to a small handful of students. And that it’s impossible to relate personally to a man who consistently keeps himself at a third-person distance. And that women figure only briefly in his narratives.^[4] And that he certainly boasts about an awful lot of war crimes.^[5] Forget it all! Caesar’s vocabulary is the easiest out of any Roman prose author and his sentences are straightforward!

Yet, however clear Caesar’s prose may be, reading him so often proves to be a frustratingly dull experience. One student writing on the Sententiae Antiquae blog was quoted as saying that Dē Bellō Gallicō “is dry and does a tremendous job detailing the frequently boring military exploits of the Roman army in Gaul. For every chapter explaining the specifics of a battle, there are five more examining how Caesar sent a dull letter to a commander that one time.”^[6] Maia Lee-Chin has often written about her journey to the classics as a Black and Chinese second-generation immigrant woman—“not your typical classicist.” She recalls the experience of reading Caesar in AP Latin: “I considered dropping the course several times, especially while translating Caesar’s De Bello Gallico; I couldn’t understand his long-winded explanations of wartime strategies, and I had no love for Roman history.”^[7] Of course, we can hardly control what students will like reading, but the odds are slimmer for Caesar than many other Roman texts. Why, then, do people insist that we should sacrifice student interest and relatability at the altar of Caesar’s perfectly accessible prose?

Repeat after me: “Latin is more than just words”

In his book Teaching Latin: Contexts, Theories, Practices, Steven Hunt recounts his experience teaching the Vēnālīcius (“Slave-Dealer”) stage of Cambridge Latin Course over the years. It used to be that he saw the story simply as an effective way of exposing students to the accusative. Not until much later did he begin to understand that there are better ways to teach students about direct objects than a story where human beings are bought and sold as property. Over time, he shifted away from a mindset where he viewed the story simply as a vehicle for Latin instruction and towards viewing the story as a story in its own right.

Hunt extends this discussion into reading Latin literature. Most Roman texts are problematic in one way or another, so it would be foolish to suggest that teachers reject them entirely. Rather, they should take the opporunity to lead students in nuanced discussions about the issues that arise from these texts. However, but with an AP-style curriculum, they have very few opportunities to actually do so:

“The choice of texts for study are a case in point. It is easy to skim through the more uncomfortable aspects of teachers’ favourite reading material or texts selected for public examination: Caesar’s De Bello Gallico is the story of what we would today describe as a crime against humanity [...]. While teachers see themselves as capable of handling such texts sensitively, time is often lacking and they often lack the experience and confidence. This begs the question whether they should be set at the school level in the first place.”^[8]

Certainly, some people would respond to Hunt by saying, “Yes, they should be set at the school level, because the simplicity of the Latin outweighs all the other concerns.” My response to those people: I’m begging you to think of the texts you assign as more than just words. The meaning behind the words is just as important—in fact, making meaning is the very purpose of language itself. How students respond to this meaning on an affective level needs to enter into the equation. Don’t fall into the trap of the AP-oriented textbook Scandite Muros, which quite offensively mines the suicide of Lucretia for sight-reading multiple-choice questions. That passage of Livy is good Latin, but not good Latin for a sight passage. Likewise, Caesar is good Latin, but he’s not Latin around which a high-school curriculum should be built.

I’m not even necessarily arguing that Pliny is the best choice. (Maia Lee-Chin is no Pliny lover either, by the way.^[9]) I do think that better options are out there from a linguistic standpoint, and besides, he’s a bit insufferable, isn’t he? But, rich and privileged as he is, he at least exhibits relatable human emotions. We’ve all felt anxiety over the well-being of a sick loved one, retrospective bafflement over our stupid adolescent decisions, the need to bring comfort to people in crisis by acting like there’s not a crisis. Pliny’s not perfect, but he’s at least open about his humanity—and you can’t get that from a guy who only refers to himself in the third person.

One more note on vocabulary

I’ll leave you with one more comment on vocabulary. Since some of the Pliny haters find his vocabulary to be a exceedingly high barrier to entry, and since he apparently uses so much less high-frequency vocab than Caesar… suppose there were an author with a higher lexical density than Pliny, the same percentage of hapax legomena as Pliny, and a smaller proportion of Top 100 words than any of the ten prose authors I analyzed? Surely, such a text would be impossible for high school students to study in AP Latin!

Surprise! These are the stats for the Aeneid AP Latin passages.

Author	Total words	Unique words	Vocab density	Hapax legomena	% hapax
Caesar	2650	799	0.302	422	15.9%
Pliny	2658	1085	0.408	678	25.5%
Vergil	2968	1244	0.419	769	61.8%

Author	% of total words in top 100	% of total words in top 300	% of total words in top 600	% of total words in top 1000
Caesar	43.4%	58.2%	68.7%	76.5%
Pliny	38.8%	54.2%	64.1%	73.0%
Vergil	31.7%	46.2%	57.6%	67.1%

Notes

1. The samples were excerpts from In Catilīnam I (Cicero), Bellum Catilīnae (Sallust), Vīta Hannibalis and Hamilcaris (Nepos), Ab Urbe Conditā Books 1 and 3 (Livy), the Cēna Trimalchiōnis from the Satyricon (Petronius), Epistulae Mōrālēs (Seneca), Agricola (Tacitus), and Vīta Caligulae (Suetonius). ↩
2. Gamberini, Federico. Stylistic Theory and Practice in the Younger Pliny. Olms-Weidmann, 1983, p. 449. ↩
3. I'm aware that the Shakespeare quote is ironic. I’m more Brutus than Mark Antony here, I suppose. ↩
4. Bellemore, Jane. “Caesar’s Gallic Woman under Siege.” Latomus, vol. 75, no. 4, 2016, pp. 888–909. JSTOR, https://www.jstor.org/stable/48620086.↩
5. Bostick, Dani. “Is It Still ‘Too Soon’ To Tell the Truth About Julius Caesar?” Ad Meliora, February 9, 2020. https://medium.com/ad-meliora/is-it-too-soon-to-tell-the-truth-about-julius-caesar-2ba5e19f25c7.
6. G., Luke. “A Response to AP Latin: A Student Perspective.” Sententiae Antiquae, September 6, 2022. https://sententiaeantiquae.com/2022/09/06/a-response-to-ap-latin-a-student-perspective/.
7. Lee-Chin, Maia. Et Cetera: An Illustrated Guide to Latin Phrases. Andrews McMeel, 2024, p. xii.
8. Hunt, Steven. Teaching Latin: Contexts, Theories, Practices. Bloomsbury, 2022, p. 170.
9. Lee-Chin, Maia. “Killing Caesar: The Newest AP Latin Curriculum.” April 3, 2023. https://maialeechin.com/ap-latin-curriculum/