You learn to read by reading, you learn to write by...

Reading.

It is, apparently, the key to everything (well, in terms of literacy, at least).

The latest issue of English Journal came the other day, and my old friend* Stephen Krashen had contributed an article.¹

Krashen's early articles were all about various language learning hypotheses, but these days he mainly writes articles that can be easily summarised by one phrase:

"Just read more books, dammit!"

His greatest and most oft cited Hypothesis was that language learners (First or Second) need language Input** - preferably input they can understand.  It seems such a simple concept, but it was quite radical in the day and still seems to slip the grasp of a lot of people involved in Education.

The idea was (and Krashen was one of the folk instrumental in promoting this) that if you want to improve your reading ability, you should read more texts and you should read more often.

This was tied into the idea that you learn to read by reading, you learn to write by writing, you learn to listen by listening and you learn to speak by speaking.

In this latest article, Krashen points out that there's research that indicates one part of that equation is a bit wrong:  you don't really learn to write by writing.  You improve your writing skills by reading more.

Writing, without corrective feedback, doesn't really do it (thank you, Stephen - I've felt this myself throughout my somewhat fruitless years studying German, but it's so nice to see it written in a citable article).

On the other hand, the more you read the more your brain absorbs how the written language works and what it should look like.

This is something I (and my colleagues) have often recommended to students as a way to improve the academic standard of their writing - that if you want to write well, you should read well written works.  It's interesting to realise it applies to writing skills across the board.

Krashen's article was, however, not actually about learning to write by reading - this was just something he was talking about (again) in order to add extra weight to the main crux of his article:  kids need more books far more than they need more tests.

The American government is busy trying to quantify education, by making things standardised and tested to the Nth degree.  Here in Australia, we're trying to do the same thing (although, it must be said, we are doing it rather badly).  Krashen would rather the powers that be put the money that would be spent on making and administering tests towards public and school libraries and better health care.

His theory is that poor literacy rates are directly related to high child poverty rates, and that you can't test poor children into becoming literate - you have to make sure they are fed, first, and then it's a good idea to give them something to read.

It's such a strangely simple yet profound idea.  So simple that any idiot paying the slightest bit of attention could probably come up with it.  So profound that it would never occur to anyone with the power to make it happen.

Go to the poorest neighbourhoods and invest heavily in local and school libraries, and make sure every kid has access to a decent breakfast and some books to read.  See what happens to literacy rates then.

What I'd like to add to this equation, though, is model readers.  It's not quite enough giving the kid a book and saying "go forth, young child, and read!"  It would be much better if they also had access to people who love books, and love to read them to kids.

Which is where Dolly Parton comes in: http://imaginationlibrary.com/

If you want kids to be more literate, start by a) giving them something to read and b) looking at what's going in in their lives that might stop them from reading.

It's not rocket science, but it's not happening.


Notes:
* Stephen Krashen is not actually my friend - I've never met the man.  I've just seen his name turn up a lot in my research.
** The Input Hypothesis.

References:

1. Krashen S. Access to books and time to read versus the common core state standards and tests. English Journal. 2013;103(2):21–29.

(You can read the article here, at Krashen's web site: http://www.sdkrashen.com/content/articles/access_to_books_and_times_to_read_versus_the_common_core.pdf)

Economics of Language Choice and Conlangs

Back in 1998 Abram de Swaan wrote a fascinating two-piece article for the journal Language Problems and Language Planning called  "A Political Sociology of the World Language System (1): The Dynamics of Language Spread" and "A Political Sociology of the World Language System (2): The Unequal Exchange of Texts".

The central argument of these articles is that people choose what second language to learn based on similar "economic" considerations that they would use to choose what telecommunications network to join (or, in modern terms, what operating system to throw themselves into).

He talked about languages having a Q-value, which is the desirability of the language based on the number of native speakers and the number of people who already speak it as a second language.  When choosing a second language to learn, most people go for one that will give them the most communicative bang for their buck, in other words.

If you lived in an area where most people spoke their own indigenous language or Dutch, you would see Dutch as being an attractive option, because it would give you a greater number of other people to communicate with.  Dutch is the center of a language constellation in that region, and has a high Q-value.  However, Dutch is but a planet circling the sun in the larger constellation where the big shiny language is German.  You might decide that German is therefore more attractive.

The biggest and shiniest language of them all is, of course, English - if for no other reason than the sheer numbers of people who speak it as a second language.  However, your personal needs in your local area might lead you to choose the big, shiny language in your smaller constellation, simply because you think you, personally, will speak to more people and have a greater supply of texts in that way.

Basically, de Swaan argues that people want consumer confidence in whatever language they choose to learn in addition to their own.

The thing that struck me about his argument, although it seemed to slip de Swaan's attention completely, is how this applies to constructed or planned languages.  He did give them a passing glance, but only to point out that artificial languages to exist, not to note that the consumer argument that he had observed for natural languages applied even more-so to those that have been constructed.

How many constructed/auxiliary languages have come and gone over the years?  The big winners in the game are the ones that look like they have legs - if you know what I mean.  Klingon has an army of geeks making sure it survives, and Esperanto has numbers.  The numbers make it more attractive to anyone who might be thinking of taking up an auxiliary language instead of, say, French or Spanish.

There was, at some point, a bit of a Betamax vs VHS thing happening between Volapük and Esperanto.  Esperanto won that consumer battle.  Now people learning an auxiliary language can choose between Esperanto, Ido and Interlingua if they want an IAL spoken by more than four people - and even then, Esperanto has the numbers.

I know one of the reasons why I'm learning Esperanto* is consumer confidence.  There are, depending on your sources, somewhere between 50,000 and 2 million speakers, and the language has been around for 125 years.  I have previously suggested that it was only mildly more useful, as a second language, than Welsh - but I've changed my mind on that point.  I have some faith that learning Esperanto is not going to be a complete waste of time.  I can't say I would feel the same about the other conlangs and IALs.

The consumer confidence element of language choice applies so very clearly to constructed and auxiliary languages it's staggering.  I'll have to see if someone has written a paper about it - I'd be very surprised if they haven't, but I have a strange feeling such a paper would probably be written in either Esperanto or Interlingua...



*I am learning Esperanto, but I'm doing it very slowly.  Don't expect posts in Esperanto any time soon.