If your goal is character recognition, then isn’t the Marilyn method quite unhelpful?
I feel like it would be better for the components of the characters to be the only (most important) cues you need to recall the pronunciation.
Marilyn method seems to rely on you “working backwards” - given the syllable, it’s easy to remember the scene, but given the character, it seems like it would be hard to remember the scene.
What do you experts think?
Parts of it make complete sense, and I find very good/coherent with the basics of well-formed mnemonic strategies. I’m glad Alex Mullen mentioned it to me many years ago.
Where I think the teaching goes astray (in the sense that people can easily get confused) is with the statement, “my system.”
I feel it would help more people if it was framed more along the lines of “these are the means and methods by which I arrived at a kind of mnemonic system.”
Also, statements like “be” combinations “never occur” could confuse people for a few reasons. Even though I only passed level III in Mandarin and haven’t travelled in China beyond Beijing and Guilin, I’ve lived with a native speaker for many years.
There are all kinds of combinations that no chart on the planet has likely ever accounted for. The fantasy that a mnemonic system is going to account for all possible pronunciations thanks to thinking about consonants and vowels in a way that makes sense for all learners is just that: Fantasy.
Correct me if I’m wrong, but no teaching is going to appear to everyone, and if you look at reviews and responses to Heisig, it is one of the most polarizing mnemonic approaches I’ve seen.
As for character recognition, I do not get the impression that this has anything to offer in that realm.
But if get something like the Outlier Linguistics add-on for the Pleco app, you can start to explore meaning and sound components with etymological tips that will accelerate your game a great deal.
The pronunciations produced by the app aren’t great, but at least offer a starting point for listening to native speakers.
If you don’t know Mandarin HQ, check them out. They have a program where you get the same phrase pronounced by multiple people from multiple areas of China.
I don’t know if it’s the same as when I was using it, but back then, you coud go through them with and without the characters on the screen and/or with and without the pinyin.
It was really great and very helpful – and a means of validating for yourself my suggestion that using tables like this to work out a foolproof means of memorizing pronunciation is unlikely.
Note too some of the evidence on how hearing works, especially with tones. Rick Beato and Rick Rubin were just talking about this at the beginning of a recent interview on Rubin’s podcast.
They skipped a few points that I’ve read in other research, but the case Beato makes is compelling and likely true for a lot of us. I’ve just seen other research which suggests that preliminary linguistic audition has a cap at between 8-12 languages, not the unlimited amount that Beato suggests.
In any case, we do well to remember that phonetic tables and phonology over all are layered onto processes that are much deeper than can be displayed on a page or screen.
The Marilyn Method does a good job in many ways, but how exactly the problems of tz and ts, etc. are solved through homophonic transliteration simply have no one-size-fits-all solution.
Which is why I’m always hammering on about the Bruce Lee approach to mnemonics…
“Be water, my friend.”
I’m biased, because I invented the Marilyn method. I created it specifically for the problem of learning the pronunciation of a character. I would agree that it is not useful for the problem of recognition. For this, I use basically Heisig’s method. But Heisig doesn’t even try to address pronunciation.
Have you ever goven thought to a mnemonic system for character recognition? It’s what I’m trying to think up now:
Damn you are a legend