Si Nous Lisions review


I finally finished Si Nous Lisions the other day, so here is my review.

This classic graded reader was published in 1930. It gradually adds vocabulary as the story progresses.
The first 5 chapters are a bit dull, being the usual themes of the classroom, families and French tourism. From chapter 6, once you have encountered about 130 words of vocabulary, there are stories, starting with the three bears, then proceeding to various interesting stories, including “la pièce de cinq francs”, which I quite enjoyed, since it matches my personal quirks.
By the end of the book you have a vocabulary of about 500 words.
If you already have at least the top 100 frequent words in French under your belt, I suggest skipping or skimming the first few chapters and starting with Chapter 6.

In terms of the amount of text, there are 15 chapters, each containing a story, plus exercises. The first page of the first chapter has ~56 words on it, Chapter 2’s first page has 88 words, and from Chapter 3 onwards there are 5-13 mostly full  pages of text with one or two illustrations per chapter. Full pages have about 300 words, so quite a bit more reading than a typical CIDEB or Easy Reader offering, which at the A1 level, can often be read within a half hour. The large amount of text doesn’t need to be daunting though, as each chapter is self-contained as a story.

Overall I rate it as “enjoyable”. Some of the later chapters I wouldn’t mind reading again in the future, if short of reading material, but I wouldn’t call any of the stories favourites.

Next up I’ll finish reading Contes Dramatiques by Hills-Dondo.


Readability Zones


I’ve just been updating my database of French readers and observing the types of books or stories in the different ranges of my current preferred readability measure.

Scores under 4 are ridiculously easy for people with an English speaking background. Currently this consists only of episodes 1 and 2 of my Gnomeville comics. Sentences are short and vocabulary is highly constrained, exploiting French-English cognates.

Scores in the 4-4.99 range are very easy: Bonjour Luc, A First French Reader by Whitmarsh, and Histoires pour les grands. They tend to be conversation-based.

Scores in 5-5.99 tend to be the short illustrated graded readers such as Bibliobus, as well as La Spiga’s Zazar for grands débutants (target vocabulary of 150). Gnomeville Episode 3 sits here due to having longer sentences compared to the first two episodes.

Scores in 6-6.99 tend to have longer sentences, including some classic graded readers such as Si nous lisions and Contes Dramatiques, as well as the 300 word vocabulary Teen Reader Catastrophe au Camping des Roses.

Scores 7-7.99 also have the more text-like graded readers, including Sept-d’un-Coup by Otto Bond, which tends to have long sentences but well-controlled vocabulary.

In the 8-8.99 range I find the first story for native speaking children, as well as more graded readers, including one with a target vocabulary of 1000 words.

The first books for adult native speakers occur with scores between 10 and 12.

Looking at the stories in the list, my own level seems to be from 7 to 10, suggesting I should continue reading more challenging graded readers in addition to stories written for French children. That is pretty much what I have been doing for a while, as well as incidental reading on the web and elsewhere.

A quick look at the relationship between stated vocabulary sizes and the 95 percentile that I have been using indicates that the required vocabulary is  roughly 1.5x  + 2600. However, I am using a token-based vocabulary whereas most would use a word family one. If I assume token vocabulary sizes are 5 times word family sizes, then the equivalence point for this model is when the vocabulary is about 770, meaning that the vocabulary load will be excessive for stated vocabulary sizes less than 770 but be ok for sizes greater than 770. That’s reasonably reassuring. Mind you this is an extremely rough estimate.

This work was based on about 100 words from the start of the text of 40 stories, but it does seem to sort things fairly usefully. The outlier based on my experience of reading the stories is Aventure en Normandie, with a score of 9.49. I don’t recall it being a difficult read.

Meanwhile I am making more progress on Episode 3 of my comic book. I decided to divide one page into three pages, as it had a lot of text and too many new language concepts for a single page. So Episode 3 will probably be 32 pages long, breaking the standard Gnomeville pattern of 28 page episodes. Hopefully it will be ready within a month.

Ford & Hicks’s Elementary New French Reader – a review


Ford & Hicks published this reader back in 1939, with the intention of making an easier graded reader than their other book, “A New French Reader”, by using present tense to start with, and introducing the other verb tenses in later stories.

The stories are mostly interesting, with a couple sourced from books I have not encountered before (“Deux jeunes aviateurs” and “Le secret du château”). It also includes Cosette, which had me sighing “not again”, since every publisher does stories from Les Misérables. However, Ford & Hicks may have been one of the first to make a simplified reader from it, so I shouldn’t grumble. What I will grumble about is the extract from Comte de Monte Cristo. You may recall in my review of Otto Bond’s series, that the escape episode of this was a highlight, and made me want to read more.

In the Ford & Hicks version, we get the initial backstory in English before reading the scene that led to the unjust incarceration of Dantès. The story then includes more English breaks between sections of French, and given the Otto Bond version, I don’t see the purpose of these interruptions to immersive reading. The Ford & Hicks version covers more of the story, and at the end summarises the remaining plot. Then they end the summary with the statement “The interest of the story weakens after the discovery of the treasure”. Unlike Bond’s version, which had me wanting to read on, this version annoyed me with the English interruptions and further annoyed me by taking away my interest in reading more of the story.

Recent trends in language acquisition research have focused on “translanguaging”, which seems to be what I’ve done most of my life, which is mixing languages together in order to keep communicating fluently with my level of knowledge (of Dutch). There seem to be benefits to doing so, but I’m starting to think it might not be so good for reading. This is the opposite of my previous thoughts on the topic, where I supervised the building of prototype bilingual ereaders that present foreign language stories, having the most difficult parts presented in the native language of the person reading it.

What may be more beneficial is reading foreign language text that resembles the native language of the speaker. That is, using cognates, simple non-idiomatic forms of expression and sentence structures that are not too unusual. This is what happens with stories in French written for native English speakers such as my Gnomeville series (Episode 1 and Episode 2 currently available), and the books by Otto Bond, and Ford & Hicks. It also happens to a large extent in stories by non-native writers. For example, I’ve seen some English graded readers written by a Chinese author, where the English is very Chinese in style, so would not be classed as “good English”. However, as long as it is known to not be “English” English, it is probably helpful to start with for people with a Chinese-speaking background. Then there should be a transition to more English-like English at a later stage. I’m somewhat more forgiving of this idea currently than I was previously.

A few more French graded reader book stats


Since my last graded reader update I’ve looked at a few more books, some of which are “classics”, in the sense they were from the “direct reading” era of the first half of the twentieth century, following the influence of Michael West’s constrained vocabulary for language teaching, the various word and idiom frequency lists created at the time, and the idea of readability. Some of these books I had already acquired earlier; but through reading some papers published at that time, I was able to compile a shopping list of other books written according to the same philosophy.

As a result, I have a new winner in terms of expected vocabulary size at the 95% threshold of reading comfort. A New French Reader by Ford and Hicks received a 95% vocabulary size of 3532, and Otto Bond’s Sept-d’un-Coup was a close second, with 3650. Bond’s book starts with a much smaller initial assumed vocabulary (97 words) than the Ford and Hicks book (523), so Bond’s book may be a better first read despite the slightly higher vocabulary score here. As seen in my first post on expected vocabulary size for 95% coverage, these are much higher scores than my Gnomeville comics, as my comics take readability criteria to the extreme.

So based on the current stats available on vocabulary, I recommend the following first graded readers for English speakers learning French.

For 6-9 year olds: Bonjour Berthe.

For 10+: Gnomeville

For adults who don’t like fantasy comics: Sept-d’un-Coup by Otto Bond – though I think there are some errors in it, and it’s out of print (and it probably counts as fantasy…).

Stay tuned for further updates.


Nabokov’s Favourite Word is Mauve – my kind of book


I recently noticed an article in an issue of Reader’s Digest while in a waiting room. It was discussing the vocabulary of Green Eggs and Ham, and other statistical aspects of writing. It was an extract of Ben Blatt’s book “Nabokov’s Favourite Word is Mauve”. Since the article was about a lot of the things that I work on, I felt compelled to get the book, and immediately read it.

The book compares a set of literary classics, best sellers and fan fiction via quantitative statistics. The author used my old friend Natural Language ToolKit (NLTK) for Part of Speech tagging etc. for the text analysis.

The first chapter looks at adverbs, which are frowned upon in writing. The next looks at statistical differences in writing by each gender. Chapter 3 discusses the history of authorship attribution, as played out with the Federalist Papers. The relative frequency of different function words tends to be like a fingerprint for an author’s writing. He uses this on co-authored works to figure out who wrote what. Next up is seeing if authors follow their own advice on writing. Then comes the chapter I saw in Reader’s Digest, that discusses Dr Seuss and readability measures. In the same chapter he notes that the average reading age of bestsellers is decreasing, from grade 8 in the 1960’s to grade 6.

Next up is a comparison of UK and US writing, including an interesting comparison on the loudness of Americans compared to English people. Chapter 7 looks at clichés – another area I have researched, albeit in lyrics instead of novels. The remaining chapters look at book covers, first sentences, and text generation.

This book did a lot of things that are closely related to my own tinkerings, as well as some of my published research. The author is a journalist and statistician, and several of the chapters, if written in an academic rather than journalistic manner, would have made good quantitative linguistics papers, with the amount of research within the book possibly being enough for a PhD in the field.

I only had one gripe about the book, and that is one sentence in which he assumes his reader knows nothing about statistics and says so. It’s one thing to explain something clearly starting from scratch or to state that something is left to an appendix for those interested. It’s quite another to tell the reader that they don’t know enough statistics. While some readers may not be insulted (eg. the “for dummies” series was popular), I usually am. I may not be the world’s expert on statistics, but I have a working knowledge, and I am capable of learning. I had a similar experience when I wanted to learn about the premise behind the Zone diet, and read a book on it, only to be constantly told by the authors that I’m fat (I’m not). The moral is: explain things clearly, but don’t insult the audience.

All in all, this was a book that’s exactly the kind of thing that I enjoy.


Vocabulary Needed for 95% Coverage


I’ve been tinkering with ways of comparing different easy readers for language learners. Previous posts I’ve used a type-token ratio or vocabulary density, which gives some idea of how likely it is you might learn new words through repetition from a text. But for something to be readable, the general consensus is that you need to know at least 95% of the words that you read. This is a level that allows people to guess the meaning of the words they don’t know.

So something I’ve been messing with recently is predicting the general vocabulary size needed for different beginner stories in French, assuming people know all cognates and all proper nouns. I’ve only been working with short samples of text so far, and there are many other assumptions and issues that make it not a perfect comparison – including bugs in my code…

Given a small set of extracts, and assuming you don’t learn the words via their introduction one at a time, as in my comic books, we have the following:

Title Vocab Size
Gnomeville Episode 1 25
Gnomeville Episode 2 25
Gnomeville Episode 3 40
Bonjour Berthe 4179
Easy French Reader 5008
Martine a la Ferme 11854
Bonjour Luc 6163

Note that this vocabulary size assumes that each conjugation of verbs is a separate vocabulary item, as are plurals etc. so will be much larger than word family figures normally used.

You can see that the one text written for native French speaking children (Martine) has a much richer vocabulary than the texts written for language learners. The figures for these look worse than they seem, because there are many words that are typically taught early to allow conversation, but which feature much lower on word frequency lists. For example, “maman” was at rank 6163 in my list. In contrast, my Gnomeville comics are designed to prioritise frequent words and cognates to optimally improve reading, at the expense of conversation. Hence the very small vocabulary sizes required.

Recently I’ve been reading a 1939 paper by Tharp that looked at measuring vocabulary difficulty. He appears to have had similar ideas about measuring vocabulary load based on the general frequency of the words, as well as a measure of density of difficulty words. I also recently acquired yet another very early graded reader, “Si nous lisions”, from 1930, which attempted to introduce new words every ~60 running words, in the style of Michael West, who seems to have been the first to use the approach. However, I have a graded reader published in 1909 in my collection, which was intended for “rapid reading”, and was part of a series that  commenced with short easy texts. I’m not sure if they methodically introduced words at specific intervals as was done by West and others following his example.

In searches on-line, I found a French adapted reader from 1790, so we’ve been at it for quite a while. I’d like to say we know more about how to write graded readers these days, but I think West had it fairly right. The only thing we can do now is make them more interesting and relevant.

Here’s one from 1800 published for those with a German background. There seem to be quite a few published in the 1800s.

Anyway, I’ll finish off here with the usual things: we need 95% coverage to read comfortably (on average). To do that with native texts requires quite a large vocabulary. But vocabulary increases as you read more. So we should read as much as possible at the level that is right for us and of reading material that interests and motivates us. My Gnomeville comics are ideal first readers in French for those with an English language background and a good vocabulary in English. The Berthe and Luc et Sophie series are reasonable alternatives for children that are possibly too young for Gnomeville, as are the ELI A0 series. Until next time…


Extensive Reading Musings


I’ve been reading some more research on extensive reading and readability lately. One paper showed gains in reading rate, vocabulary and comprehension with students reading about 150K words over 15 weeks at an intermediate level. This was contrasted with another study where learners read ~65K words over 28 weeks and failed to show improvement. I think there is probably a threshold of some kind where you need to read a certain amount per week to improve language skill. The amount probably varies with the level of skill you already have. Someone still improving their knowledge of the most frequent 400 words of the language will not need to read as much to achieve vocabulary gain (assuming appropriate graded readers) as someone reading at the 2000 word level. The study that showed gains had students reading with vocabularies of 800+.

Given the 10K words per week guide, and the typical reading rate in foreign languages often being around 150 words per minute, that equates to about an hour of reading per week, or 10 minutes a day. That’s not a bad aim for maintaining and hopefully improving your language skills.