I used to seriously question the Comprehensible Input method for learning ancient languages, but now I strongly recommend it as fully authentic and the best method for gaining reading fluency and becoming a lifelong learner of ancient languages.

However, I still cannot bring myself to recommend the Lingua Latina Per Se Illustrata textbook (LLPSI) as a standalone Latin course. I understand that it works. And I am glad that many motivated people have acquired Latin through it. However, I cannot bring myself to like it, even when I try as hard as I can.

Let me start with how I had formerly come to reject Comprehensible Input together with LLPSI. I learned Latin through a reading method with the Cambridge Latin Course. In practice, my teachers taught me Latin through a mixture of reading pleasurable stories, getting us to orally perform the stories, and drilling our grammar with grammar-translation exercises. I was initially intrigued by LLPSI because I had heard it was a more intuitive and reading-based course than the one I had learnt from, and I believed reading – any kind of reading – was inherently more enjoyable than grammar-translation.

However, when I read LLPSI myself, or used it as a resource for Latin tutoring, both I and my students found it very hard to endure. I couldn’t put my finger on it, but something seemed very unpleasant about the stories. The narrator seemed to dwell on bodily harm in chapter 3 when he wrote the onomatopoeia “tuxtax, tuxtax, tuxtax…” to the sound of a child being beaten (the illustration showed a man with a face wrinkled with anger, hitting the boy on his lap with a stick). The incident was neither funny nor tragic. The boy’s torment simply happened. The tone seemed to lack human empathy, but presented all things, whether good or evil, in clinical detail. Frequently the stories did not follow the conventions of narrative, of motivated characters driving a plotline (such as the stories in the CLC), but instead were written like information dumps, with the story being merely a framing device for saying more words, more horrible words. My dislike of the book’s construction hardened; it grew into a burning hatred the more I wanted to like it, and the more I read of it (except for the chapter on pirates – that was actually very interesting. There actually were a couple chapters in the middle that weren’t too bad).

Meanwhile, I was tutoring a lot of students from a nearby school who had significant gaps in their Latin learning because for a whole year in middle school, a Latin teacher there had taught them from LLPSI. Maybe this wasn’t LLPSI’s fault – perhaps the teacher or the students were sloppy, or the class was noisy, or it was because of the sudden change of Latin courses. Who knows? But I certainly did not hear a ringing endorsement of the LLPSI course from the boys I tutored. They said they were told to “work out the general meaning of the sentence from context” when reading LLPSI, and be content with not having every word exactly right (as I later learned, this is not how CI should be implemented). I asked them to interpret sentences from LLPSI aloud for me and they were getting every second word wrong. I drilled them on grammar to fill their gaps, and they learned the rest of their Latin through a mixture of reading and grammar-translation methods – the Oxford Latin Course.

As this was happening, I was reading only ringing endorsements of LLPSI on the internet from people who supported Comprehensible Input as a learning method. I couldn’t reconcile this with my experience. I came to the conclusion that these internet reviewers were extremely motivated adults who didn’t mind reading stories about boys being whipped, or info-dumps spelling out things like “it is necessary to breathe to be alive”. These self-learners could not access the Cambridge Latin Course or Oxford Latin Course due to the steep price (at roughly $60 per book, the OLC would set you back at $180, while $50 a book for the CLC makes it $250). Their only affordable alternative to LLPSI was the driest of all textbooks – the extremely dense grammar-exercise courses like Wheelock’s Latin, Jenny’s Latin, So You Really Want to Learn Latin, Latin for Dummies, and so on. Given that choice, I thought, it’s no wonder they supported LLPSI. I concluded that all the twaddle about Comprehensible Input was people justifying to themselves why they should keep persisting with LLPSI and making themselves feel like they are accomplishing more than other Latin learners, as a motivation tool to get through LLPSI.

Fast forward to this year. I recently came across a very well executed Comprehensible Input (CI) ancient language course – Aleph with Beth. This is a biblical Hebrew course taught entirely in spoken Hebrew, with props, gestures, and stock footage of animals, people, and natural scenery. The videos are very simple in some ways and very complex in others. They are not structured as stories, but as an instructor telling you joyfully about the things that exist. As I watch these videos, I feel like I am discovering the Creation afresh. She holds up a realistic miniature horse figurine and says “SUS!”, and now I know, that is indeed a “sus”. She points at two horses together and says “SUSIM!” and I know, yes, they are “susim”. And then I am hooked to the screen. I am mimicking what she is saying. I am asking and answering her questions out loud. I am listening and echoing and acquiring the Hebrew language, and just by paying attention, I am 99% confident of the meaning of the words she says.

This is going to sound strange, but I never considered myself a “gifted with languages” person or a “language lover”. I liked Latin specifically, and then Greek specifically, and then got into self-teaching Hebrew through a grammar method. Each time, there was a lot of hard work and joy mixed together, but it was worth the hard work because I then got to read ancient texts in their original languages. Most of the time the real payoff for my efforts would come months or years down the line.

Aleph with Beth is inherently engaging from the start. It does not feel like work, and it is not boring either. It respects you and doesn’t condescend, but it also takes as much time and joyful repetition as you need to learn each word and structure one at a time. After watching the videos my brain feels tired, but while watching them, I feel energised.

I have been inspired by Aleph with Beth to work CI strategies into my Latin teaching practice, with very pleasing results. I might share these strategies in more detail in the next few blog posts, but in summary, I have been showing CI Latin videos from ScorpioMartianus that were adaptations of LLPSI. I have also implemented the 7-step-Ranieri-rereading method with a variety of year levels, from total beginners in year 7 to our final year classes in year 12 studying Vergil’s Aeneid. I have made a vocabulary flash card set for my year 9 class with only pictures and prompting questions (eg. a picture of a book with the question, “quid est?”) so that students see the picture and have to orally answer the question by saying the target vocabulary word. I have been rewording grammar explanations to be more about using Latin phrases to explain other Latin phrases. All of these things have been immediately showing good results – students internalising and recalling the vocabulary, reading prepared stories at the speed of speech, understanding even complex grammatical constructs without first having to force them into an English pidgin.

I was so happy with the results of CI strategies that I felt that I should pick up LLPSI again and see if it actually makes sense as a course now. I liked the LLPSI adaptation videos by ScorpioMartianus. I liked Aleph with Beth. I respect Luke Ranieri as a very capable instructor, and he recommends LLPSI as the best Latin course ever made. Surely now I can actually put aside my prejudices and finally like LLPSI.

I read a few chapters again and think “Dear God, I really hate LLPSI. What is wrong with me?”

The position I am left in is this: I love Comprehensible Input, and I can’t stand LLPSI.

And now I think I know why.

When you watch Aleph with Beth, Beth sits in a clean and bright room. She holds one prop in front of her. There are no distractions. She shows you the concept – a horse. She asks what it is. Then she says it is a “sus”. She repeats “sus” a couple more times, gesturing to this new discovery. Later in the lesson, a video (or two, or three) of a horse galloping across a field flashes brightly on the screen. After our eyes register this, she says “sus” again, but I am already starting to say “sus”. She asks her assistant – her husband Abram – to show her “sus” from among several other animal models. He points at it and is praised, or else he points at the wrong animal and is gently corrected.

In Aleph with Beth, you are shown the concept before learning the word for it. You see a horse and then are told it is “sus”. You are shown 3 or more images of the same thing. The props are lifted up, and you see her tangibly interact with them in three dimensions. You have a very vivid image of the thing that is being talked about.

When I read LLPSI, it feels like I’m seeing about 20 new words per paragraph. Some of these have pictures somewhere else on the page. But what predominates is the text itself. You are supposed to read the text first and then use it and your intuition to work out the meanings of the unknown words and structures. Some of the glossed definitions are written in other Latin words. Sentences with new vocabulary are just as long as sentences with old vocabulary.

The problem here is that LLPSI tells you the word first, which at first is nothing but a set of letters, and then you have to exert effort to find the meaning. Aleph with Beth shows you the actual thing first, and only then do you learn its name. In Aleph with Beth, the meaning comes before the language.

I find that this mimics the process of a child acquiring terms in their first language, especially in what I have seen of the Montessori style of education. First the child is shown a set of wooden cubes arranged according to size. They pick up a bigger one and feel its greater weight, then pick up a smaller one and feel its lightness. Only after that experience are they eventually taught the words “big”, and “small”, gradually moving to “bigger”, “biggest”, “smaller”, “smallest.” The meaning precedes the terminology.

A second practical problem I find with LLPSI is the printed format of its question-answer dialogues inside of the main chapter stories. For example, a paragraph could consist of the following sentences printed together:

Num Crēta oppidum est? Crēta oppidum nōn est! Quid est Crēta? Crēta īnsula est. Num Sparta īnsula est? Sparta nōn est īnsula! Quid est Sparta? Sparta oppidum est. Rhēnus quid est? Rhēnus est magnus fluvius. Num ōceanus Atlanticus parvus est? nōn parvus, sed magnus est oceanus.

In a video, these questions and answers work well, because you don’t see the answer right next to the question. You have a moment, however brief, to interpret the question and start to say the answer, before the narrator actually gives the answer. But when you read this, the answer is already there. I find I skim through these written questions and answers very quickly and without much motivation, but when the same dialogue is presented in a video format, I try to get the answer before the narrator can. I find a similar thing with my students when I show them CI videos: a number of them will try to say the answer out loud ahead of the narration. It’s a thrill to beat the narrator. It is no joy whatsoever to have to read an answer printed inline next to a question.

I feel like the video adaptations of LLPSI have achieved what LLPSI in itself has struggled to do. The course cannot be blamed so heavily, as it is a remarkable thing to have held up over 50 years. I think the greatest problem is the physical limitations of a book format: there is a limit to how many repetitions of words and pictures can fit in a printed book without it starting to seem wasteful to the environment, but a video series can take its time on repetitions of minute details one by one, and show a thousand times as many images in the same period of time.

For those who are currently using the LLPSI, I would personally recommend listening to an audiobook of the passages (anything people have read aloud of it on youtube or elsewhere) and would encourage you not to be ashamed to look up a translation of LLPSI in your native language to ease the process of initially comprehending the text, if necessary. Re-reading the Latin while re-listening to LLPSI passages would certainly be a good reinforcement for the target language features, even if the meaning was initially found through your first language.

I greatly look forward to seeing more comprehensible Latin videos being made, and I would also like contribute to this effort by making some CI videos of my own. I have seen the results of well-implement CI methods and I cannot endorse them enough.