AP Latin: There’s Bad News…And…Worse News

**Updated 2.25.21 with details from this post**

I ran texts from the AP Latin syllabus through Voyant Tools:

  • 6,300 total words in length
  • 2,800 forms (i.e. aberant + abest = 2)
  • 1,100 meanings/lemmas (i.e. aberant + abest = 1 meaning of “awayness”)*

Based on the research of Paul Nation (2000), 98% of vocabulary must be known in order to just…read…a text. According to Nation’s research, then, Latin students must know about 6,175 words they encounter in the text in order to read the AP syllabus texts. That’s a text written with 1,100 words. To put that into perspective, it’s been reported that students reasonably acquire ~175 Latin words per year, for a total of something more like 750 by the end of four high school years. Needless to say, there’s a low chance that all 750 would be included the Latin on the AP, and that varies from learner to learner. Even if they were, though, 750 is still only 68% of the vocabulary at best. Although this percentage isn’t the same as text coverage since it doesn’t account for how many of the 1,100 words repeat, it’s safe to say that the number isn’t going to be wildly higher. Even approaching 80% text coverage is not good. We know that reading starts to get very cumbersome below 80%. This is just one reason why no student can actually read AP Latin. Oh wait

****Those figures are just for Caesar****

I ran the Virgil passages through Voyant tools as well:

  • 5,400 total words in length
  • 3,200 forms
  • 1,500 meanings/lemmas

Despite being shorter in length, Virgil uses 14% more forms and 30% more meanings/lemmas. Since students read both authors, though, I combined everything into Voyant Tools once again to see what students are really expected to read and understand:

  • 11,700 total words in length
  • 5,700 forms
  • 2,600 total meanings/lemmas
  • 370 meanings/lemmas in common between both authors

That’s a lot of Latin, a lot of forms, and a lot of words (and only 14% in common between the two authors). Again, based on Paul Nation’s research, students must understand 11,450 of the words they encounter to be able read the AP Latin syllabus. And that represents texts with 2,600 words.** With about 750 words acquired by the end of year four, that would be just 30% of all the vocab in Caesar and Virgil at best. This is exceptionally low. N.B. AP syllabus reading begins at the start of the fourth year. By that time, students likely know just 600 words, which is 24% of the vocab!

Yeah, but AP stands for Advanced Placement.”
It could be argued that AP courses are intended for the academic elite, and not “typical” learners. Assuming an “advanced learner” can acquire twice as many words, though, which is a biiiiiiiiiig assumption, that still only results in 60% of the vocab. Could that vocab account for a text coverage of 80% of the 11,700 total words? If so, that’s still nowhere near a readable level. Clearly, students aren’t actually reading Latin on the AP syllabus. If not reading Latin, though, what is being measured on exam itself?

Before that’s addressed, it should be noted that the AP Latin syllabus actually has required readings in English, too. These are translations of the Caesar and Virgil passages in Latin, only entire chapters, not just excerpts. That preparation amounts to 32,300 total English words in length, which not only is translated Latin literature written at the Distinguished Level, but also unfamiliar references from the distant past. That is, it’s been observed that a test on the content of Caesar and Virgil, entirely in English, would be quite challenging on its own, and that’s in a language students know well. But this analysis has been on the AP syllabus reading list. Considering the scope, there should be a decent amount of Latin on the AP Latin exam, right? It turns out there just isn’t that much on the exam.

The Exam
As for the exam itself, students are assessed on a variety of skills and knowledge. However, actually reading Latin is a very small part of the exam. In the sample exam, there were perhaps five actual reading comprehension questions in the first section, and another four in the second section. Given other online samples and reports from various teachers, the actual Latin passages amount to something like 600 total words in length. To put that into perspective, that’s 5% of the AP syllabus in Latin, and if you include the required English translations, that’s close to 1% of the total course content. Note, too, how this analysis is simply addressing comprehension—step zero—let alone additional cultural competence needed to compare, contrast, and analyze the Latin texts.

So, for years now, I’ve heard how the Latin on the AP exam is written at the same Distinguished level as the English on its own Lit & Comp AP exam. That’s enough bad news. However, I had no idea that the scope of the AP syllabus was beyond research and reason, and ultimately doesn’t really focus on the supposed universal goal of teachers: reading Latin. At an anecdotal level, this analysis certainly confirms stories from teachers moving towards a more comprehension-based and communicative language teaching (CCLT) approach who also offer AP Latin. Such teachers report that despite increased enrollment filled with students reading more Latin at higher levels of comprehension, they aren’t necessarily doing well on the AP Latin exam. That is, by all other accounts, these Latin programs are thriving, but that success isn’t represented by AP exam scores. Of course, it’s clearer now that the AP Latin exam doesn’t really measure reading comprehension (for that, try ALIRA, or just hold a discussion).

But The College Board has built its machine, convincing universities and high schools that the test scores are valid measurements of valued skills. There’s a lot at stake here, too. People have published support materials for exam prep, which means there’s money. Big money. Districts pay for teacher training, and for students to take the exams. Big money. Perhaps the most disappointing result is backwards designing a Latin program’s curricula to begin teaching AP content an entire year earlier due to the scope, as well as denying a Latin experience when AP is the capstone (here, meaning the final course offered in high school, i.e. lack of a non-AP track). This, essentially, amounts to “weeding out” typical students despite what research says about realistic expectations of appropriate reading level.

So, working around that AP machine, or even tearing it down, is tricky. Like I wrote in Why Your Language Teacher Failed You, for now we’ll have to wait until teachers challenge the status quo. Not a very easy thing to do.

*This is the lowest possible figure, and best scenario. The reality is that many, many forms of
esse, quisque, and other pronouns appear throughout. Words with different stems can be less-recognizable to complyetly unrecognizable (e.g. est vs. fuērunt), and likely should be considered two different meanings. However, for consistency, these were counted as one (1).

**The reality is probably much worse. For example, a student who can’t recognize all stems and inflections starts moving towards not understanding the combined figure of 5,700 different word forms.

***I reviewed thousands of word forms—6,124 in fact. I’m sure I miscounted at some point, and I’m sure there were typos in the original Latin texts I used. For this analysis, then, I actually rounded almost every figure down to the nearest 100! Still, even with potentially hundreds of mistakes not accounted for, the situation remains grim. What you’re looking at is a very forgiving and favorable outcome of what’s probably much worse.

12 thoughts on “AP Latin: There’s Bad News…And…Worse News

  1. Lance, this is an excellent set of stats to consider, and quite frankly, it’s the pretty damning evidence that most of us have suspected concerning AP Latin for a long time. And you do not even address the issue that even if one had the requisite 98% vocab, half of the readings on the syllabus are in poetic form–considered the most distinguished poem in all of extant Latin literature. Even native speakers of English who have nearly 100% of the vocabulary find poetry difficult to read and understand. So, even worse news on top of it all. Thank you for taking the time to gather this data and reflect on it with us.

  2. I would also add that this is a test were you can be taught the tricks to increase your score and ‘beat’ the test, rather than having to rely entirely on your knowledge of reading Latin…

  3. Lance, thank you so much for putting what we’ve known into these words and data. I’m very clear with my AP students regarding this information, and while I myself never took the AP Latin exam, I do applaud my students who choose this route because of the intensity of the curriculum. I do warn them that the AP Latin syllabus is for a course ABOUT Latin, not simply Latin. Until recently, the AP Latin course was our “capstone” as well, which I felt was inadequate for our students. I really wanted to offer a fourth year in our Spoken Latin curricula, and thanks to our Academic Affairs office, that is now an option and has been utilized by two students this year (the first year it is offered). I’m excited for the coming years where students know this fourth/non-AP year option to exist and they can take advantage of it as well! Thank you again so much for this data and narrative. I greatly appreciate reading it!

  4. I have long thought that the answer lies in changing the nature of the test as well. But as you say, money is king. What I would lobby most strenuously for is the creation of a new test, “Latin Language Skills” or the like. Now it is an “Advanced Latin Literature” test. Think, literacy vs. Milton/Shakespeare test. I believe this would bring us into line with modern languages. To sell it we would have to convince AP that more kids would take this test and thus more money for them. I believe this is true. Parents, students, school districts all want AP courses. Colleges want to lure kids in by awarding credits for them. But the percentage of students who could gain appropriate credit in a Latin Language test will be much higher than for the current exam. Hence, or so I surmise, such a change would also be a plus for the consumers of the test. The trick is to start a discussion with AP folk who make and grade the test, perhaps starting with a panel at ACL

  5. I’d be interested in where you are deriving “it’s been reported that students reasonably acquire ~175 Latin words per year” from. Not that I’m necessarily disagreeing, I’m curious about the basis for this number.

    • To follow up with updates, late last summer I actually did hear of someone expecting students “knowing 250-500 words after Latin 1.” That’s the highest figure I’ve ever heard, though cannot say that it was from a comprehension-based teacher (vs. some other kind of memorization-based teaching). For the sake of conjecture, let’s say they *were* comprehension-based…

      Were momentum to continue and student vocab end up at something like 1270 by the start of that AP course…that’s still a text coverage of 50%, and woefully inadequate.

  6. Lance,

    I’ve taught the AP curriculum a couple of times with students who learned via the grammar-translation method. It is a bit fast-paced, but there are a few mitigating factors: 1) the students will have read the entirety of the AP Latin selections at least twice (often three times) before taking the AP test, 2) rare or unusual words are often glossed, 3) the translation portion is chunked so that you cannot get points by merely getting a form or definition right, you have to get the entire phrase right. This means that while you may get 50% of the chunks correct and still get a 5, there is a level of understanding that is necessary here, since the 50% you got wrong may be wrong in just minor ways. This also makes up for the fact that students may not know all the vocabulary. 4) Multiple choice for sight passages contain extensive vocabulary aids for words not expected to be known. At some point they will need to learn how to read Latin that they do not know all the vocabulary for. I suppose the question is: is the fourth year of Latin the right time?

    That said, I do think the AP syllabus is a bit too long for most students; nonetheless, my best students have done a fine job reading the Latin, engaging with the text, and and preparing for the test. The AP test should not, in my opinion, be the capstone for any Latin sequence intended for all, since most students will not do well in it – especially when in a CI oriented approach.

    Thank you for your link to the ACTFL. It looks awesome!

    – Eddie Kotynski

    • THanks for the reply. However, my post isn’t about the students who do well on the exam. It’s about everyone else. I’m more concerned with your comments regarding CI. You’ve contrasted it with grammar-translation, yet any student who learns Latin via grammar-translation receives CI to an extent. Still, you also mention in point #3 that identification of forms isn’t really part of the exam, anyway, so it sounds like any additional grammar knowledge isn’t necessary on the AP exam. What is it about grammar-translation, then, that you see preparing students for the AP exam?

      But about mitigating the vocab, if your point is that the vocab both 1) learned by the typical student is higher, and 2) the burden is much lower given glosses, I cannot see the result changing by much. We’re still talking about a text coverage of what…50%…60%…70%? That’s too low for reading according to the research, so whatever your students are doing with the Latin, it’s something else. I’m not sure if you saw this, but my take on it all is that success with grammar-translation, and/or AP exam is almost completely out of the teacher’s control. It’s just a few kids each year with excellent memory: https://magisterp.com/2020/03/23/grammar-translation-not-really-a-method-resisting-it-now-more-than-ever/

      • Thanks for responding, Lance. I agree with much of what you say. I am not defending GT, nor do I think that those who employ it are truly reading. It is something else which has its uses but is not the ideal way to learn a language. Nonetheless, I don’t think that most students will be able to access the Aeneid ever in Latin in their lifetime if they have to know anything like 98% of the forms and vocabulary. A lot of the choice, then, between pure CI and a mix of GT, CI, and various other approaches depends on your goals for students. Whether any specific goals are worth pursuing is another question.

        I really appreciate what I have learned about language-learning from you and others and hope to continue to learn from your work.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.