What's wrong with LLPSI, Part 1
(This post co-written with Gregory Stringer)
I (Seumas) am happily on record, and will continue to say, that Ørberg's Lingua Latina per se Illustrata, and in particular volume 1, Familia Romana, is one of the best language textbooks in any language. Not just that, but it is so by a very wide margin. It demonstrates a deep and profound instantiation of the Direct Method, and succeeds very admirably at being 'per se illustrata'.
However, it's not perfect. Some of its faults are innate - a textbook cannot escape being a textbook, for instance. Others are characteristic of the text itself. In part one, we are going to look at what we think are some of its major drawbacks on the language side of the equation. In part two, we will explore some of the more content oriented problems in the book.
- Assuming that one you understand it, you know it.
Ørberg does a really good job of introducing new language features, illustrating or exemplifying them in a way that makes them understandable based on the building blocks you've had so far, and then covering a range of forms and permutations. Chapter 17 is a good example as he walks you through exemplars of all persons and number of the passive present across the conjugations. However, there is this assumption in Ørberg that knowledge is binary - you go from 'not knowing' to 'knowing', and then you know it. Which means that later on the text always assumes that you know a feature you have been taught. Humans don't learn, or know, like this. The process is far messier. Familia Romana would be a better book if it had a lot more redundancy built in.
- Grammar driven curriculum
Related to this, FR is shaped and structured around incremental acquisition of grammar: morphology and syntax. We also have good reason to believe that this isn't how humans acquire language, and so FR is still shaped by a grammar-driven curriculum. This is why I say often that LLPSI is not a 'CI-book' - it's not a book that was written based on the idea or principle of 'Comprehensible Input', or the practices that shape contemporary communicative-based language teaching. It is, or can be, comprehensible input in the sense that it is a wonderful text that many people can read and understand without much help if any, but it is not a "CI-based" method of teaching. It reflects, accurately, Ørberg's understanding of the Nature Method. As best as we can tell, Ørberg did try to apply some understanding of acquisition orders, based on other languages, to his sequencing, but there is still very much a sequencing going on.
- Vocab and volume
LLPSI has a tonne of vocab. Not too much, I don't think. I'd rather students got the 1800 or so words of FR compared to the smaller vocabulary counts in other textbook offerings. However, for the amount of vocab he presents, it should be a longer book. Or it should have companion books that introduce no new vocab or grammar, but just tell other stories (yes, I'm aware of the supplements. What if we had more, and more diverse, supplements?).
- Some grammatical features come too late
The subjunctive, in particular, appears quite late in the book, giving students inadequate space, in this volume, to attempt to assimilate and acquire its forms and usages. Similarly, students don't meet the 1st and 2nd persons until chapter 15, and tenses other than the present until 18, when they finally do come thick-and-fast. Any curriculum is going to introduce some things early, some things late (unless you choose a totally different organising principle), but these seem particularly difficult.
