Hacker News Reader

Comment by SoftTalker | Hacker News Reader

This seems so weird. When I think about how I learned to read, in the 1970s, it was (as best I can remember) first learning the letters and the sounds they make. Then starting to read words by "sounding them out." I never remember learning about "context" or "what word would make sense here" or "what do the pictures show." Pictures were just there to make the pages more fun to look at for a 7 year old.

Of course after some exposure and repetition you start to recognize whole words at a glance. That's just natural, but I never remember learning to read by memorizing whole words.

View on HN

You learn to walk before you learn to run.

This should be obvious, but a surprisingly large number of people don't get it. They don't see "running" as the logical next step after "walking", but rather as an alternative to it. "Why are you teaching my child to walk, when you could teach him/her to run instead?"

They imagine that the fastest way to get to the advanced lessons is to skip the beginner lessons. Yeah, it's a good way to get fast to the Lesson 1 in the Advanced textbook... and to remain stuck there forever, because you don't know the prerequisites.

The article describes what happens when the people who don't get it are setting the rules for others to follow.

Someone noticed that the advanced readers read fast (correct), sometimes entire sentences at once (kinda correct), and concluded that the proper way to teach children is to insist that they do it from the start (utterly insanely wrong). You should increase your reading speed naturally, as you get lots and lots of practice; not because you skip letters - that's actually when we should tell the kids to slow down and read it again.

Or maybe, listen out, not everyone is stupid and the reality is just really complicated?

As an anecdote, my daughter was learning reading in her native language in school starting with letters, then syllables and had a very hard time moving past that with a lot of support from teachers and family.

She started learning to read in English almost 5 years later by reading the whole words from the start and outperformed her reading and comprehension speed to her native language very quickly.

There are huge number of variables in play and common sense frequently doesn't work.

A small portion of people are different, but we should start with the simple way first.

And while context can get one ahead early, you don’t want to be like the adult who couldn’t actually read.

Don't know why this has to constantly be mentioned, but people who read this website, and their children, are not representative of the general population.

It is well known that some kids will learn to read no matter how they are taught. Most kids will not.

Yeah, people are different. I guess there may be some kids for whom the slow reading does not work for some reason, and who benefit from reading the whole words. But in my experience, most kids start making mistakes when they try to read too fast.

Perhaps the method was helpful to some children, and the mistake was to prescribe it to everyone.

In the 90s I was taught to read via phonics. Context was mentioned further down the road as a tool to reach for when one understands all but one word in a sentence, in which case context can be used to infer the meaning of the mystery word sometimes (but not always).

I can’t imagine not having a functional knowledge of phonics. That must make long unfamiliar words daunting and reading overall more scary than it needs to be.

>first learning the letters and the sounds they make. Then starting to read words by "sounding them out."

This is called "phonics" and was universal until recently. The 1980s had commercials advertising "Hooked on Phonics works for me." - Hooked on Phonics being a books on tape program to help children read.

TFA says phonics was popularized in the 1800s.

That's how writing used to work for the longest time. Each letter has a sound, and you write down the letters that match the sounds you make when pronouncing the word. Two people might not spell a word the same, so the only viable way to learn would be what is now apparently called phonics.

We only really started to standardize spelling in the 1500s. Which I guess means that by the 1800s English spelling and pronunciation had drifted far enough apart that phonics was a concept worth putting in words.

In most languages with alphabets the pronunciation of letters is consistent enough that the issue doesn't seem to come up a lot. Phonics is just the obvious way to do it in those cases

Arabic is very phonic. I found it pleasurable to learn a little of it. Like a lisp after using c++!

But you have to guess every vowel, correct?

> in the 1970s, it was (as best I can remember) first learning the letters and the sounds they make. Then starting to read words by "sounding them out."

USSR, 70s, the same, my older cousin, 5th grader a the time, taught me to read that way before my first grade. (It was pretty normal to learn to read before starting the school. The writing though was taught at school.)

That's because the Russian alphabet is phonetic (in one direction). So you just need to learn the sounds corresponding to the letters and a handful of rules used to combine them. After that, you can sound out the words aloud, and then it's just a matter of practice.

English is not really phonetic anymore, so this approach doesn't quite work well.

But at the same time, English teachers don't want to go the full Chinese route. Because if learning letter combinations is somehow "colonizing" ( https://time.com/6205084/phonics-science-of-reading-teachers... ), grinding through thousands of words to memorize their pronunciation is probably something like torture and genocide.

Now that you mention it, yes we did learn some combination sounds, and rules about when letters are hard, soft, or silent etc. And exceptions, such as "ph" sounding like "f" but those came later. The first books were like "Dick and Jane" with very simple words.

>English is not really phonetic anymore, so this approach doesn't quite work well.

That seems to be one of the main components of Russian accent in ESL.

Not really? The accent source is typical for any pair of languages: different sets of sounds. E.g. Russian doesn't quite have sounds for "th", "w" ("William"), "a" (as in "apple"), etc.

What do you mean by "in one direction"?

it would mean that each letter has one and only one sound, but multiple letters can share the same sound. or if it is the reverse direction for each sound you only have one letter, but multiple sounds can share the same letter. which one is true for russian i don't know.

i learned to read the cyrilic letters, but i didn't learn russian (i did try though) but with that knowledge i could read cyrilic texts aloud to someone who understands the language, assuming i learned all letters correctly and the first case is true.

in the second case i could write down anything i hear. much harder, but as a traveler that would actually be useful. be able to write down names and addresses i hear when asking someone for directions for example. i did learn to write (well, type) korean that way, but of course i had to ask a local to proofread what i wrote since i would not be able to spot mistakes.

In Russian, unstressed vowels are reduced so they are pronounced ambiguously. And when you try to write them down, you need to choose the correct letter for the full-length vowel. There are also double consonants that often are not pronounced differently.

On the other hand, if you just sound out the words syllable by syllable with full-length vowels, they will be completely understandable. You'll just sound a bit over-formal and/or robotic.

There were several attempts at spelling reforms, but only the first one (in 1917) stuck.

> English is not really phonetic anymore, so this approach doesn't quite work well.

For each letter you can find a way it is pronounced most frequently, and then take a subset of English consisting of words that follow those rules completely. (For example, the word "cat" is pronounced as a concatenation of the most frequent way to read "c", the most frequent way to read "a", and the most frequent way to read "t".) You learn to read these words. Later you start adding exceptions, for example you teach how to read "ch", and then you add the new words that follow the new rules. Etc, one rule at a time. (You leave the worst exceptions for later grades.)

>> This seems dehumanizing, this is colonizing, this is the man telling us what to do

If you feel "colonized" by reality, I guess you can rebel, but you shouldn't expect reality to reward you for doing so.

> English is not really phonetic anymore, so this approach doesn't quite work well.

I presume you mean it's not particularly 1-to-1 spelling <—> phonetic.

It is highly phonetic, but it does have alternate mappings between individual or adjacent letters and sounds. And silent letters or syllables.

But alternate rules are rarely random. There are usually many words represented by each rule. And those words often have similar overall spellings and phoneme patterns.

> English is not really phonetic anymore, so this approach doesn't quite work well.

English pronunciation <-> spelling is actually pretty predictable as long as you aren't considering letters/phonemes in isolation.

1. recognize whether it's a compound word or a word with affixes, and if so break it down (e.g. shep-herd)

2. recognize the "origin" of the word - at a minimum, "native" (German/Norse) vs "foreign" (Greek/Latin/French mostly, though others come up) is usually obvious, though sometimes it becomes necessary to be more specific or even care about when it was borrowed.

3. recognize the stress pattern in the word, and how that will affect possible vowel sounds

4. recognize the letter pattern or sound pattern (depending on which you're starting with)

These are not independent recognitions; often one or two is enough to imply everything you'd need to know about the others (and this in fact reinforces the pattern recognition humans are so good at).

An informative example is "arch". "ar" fixes the pronunciation of the "a", and "r" is not ambiguous (ever, for rhotic accents; after syllable division for non-rhotic accents). The "ch" is pronounced "tsh" for most words (whether German or French), but when it is of Greek origin (or at least came via Greek) it is pronounced "k". Usually such words are compounds with other visible Greek components.

> English pronunciation <-> spelling is actually pretty predictable as long as you aren't considering letters/phonemes in isolation.

Yeah, and you also learn the etymology of each word. With plenty of exceptions.

I learned English mostly as a written language, by reading books. And for _years_ after moving to the US, I had a problem with pronouncing words that I knew perfectly well how to spell.

E.g. I was confused when a doctor told me that I had "neumonia", even though I knew the word "pneumonia" perfectly well. Or that "gearbox" is not pronounced "jearbox".

> but when it is of Greek origin (or at least came via Greek) it is pronounced "k"

Or Latin. I volunteer to teach English to refugees, so my rule of thumb: if a word is similar to a Russian/Ukrainian word then it's pronounced with a "k" sound. But there's also a bunch of French words where "ch" is pronounced as "sh".

But really, the main rule is to just memorize what the pronunciation is.

The Russian alphabet is not phonetic. а can be pronounced а, и, ы; е can be pronounced и, ё, э, and so on, and most consonants can be pronounced in two ways depending on the vowel that follows, or the presence of ь. You need to know where the tonic accent lies in every word to be able to pronounce it, because the position of a vowel w.r.t. the accent modifies its pronunciation. It is more phonetic than English or French, but less than Belorussian or Finnish or Spanish.

Germany, 2010s: We learned the letters with pictures of animals, that started with that letter. Also complicated words were initially replaced with inline pictures.

Reply on news.ycombinator.com