Issues, Tweaks, and Examples
I’ve taken an extended break courtesy of the Step 1 Exam, but I’m excited to get back into things by wrapping up a previous blog. Just over a year ago, I wrote “Learning Chinese with Memory Techniques: Part 1 – The Gist.” I’ve waited far too long to get this second part out, but here goes. What I’d like to do is simply discuss the method in more detail—complete with examples, tweaks I made to Serge’s approach, and issues I faced. The good news is that, nearly a year in, I’m still happy with the method and will continue using it in earnest.
For anyone seriously interested in learning this way, I’d strongly recommend reading Serge Gorodish’s original description. The technique isn't that different from your garden variety memory techniques, but all credit goes to him for the inspiration.
As I discussed in Part 1, the crux of the method is that—thanks to the fact that Chinese consists solely of one syllable words with a finite number of beginnings (56) and endings (12)—we can unambiguously, systematically encode any given word. We place character x (beginning x-) into location y (ending -y) doing action z (the word’s translation), and voilà: We can now remember that xy means z. This is great, because many Chinese words sound incredibly similar to the foreign ear, rendering usual mnemonic methods less than ideal.
Here’s the system as filled in by me (with an excel version for download).
Serge breaks the 56 beginnings into 4 groups, each of which he assigns a category: men, women, comics/cartoons, and mythology. Like him, I made group #1 all guys and group #2 all girls. His latter two categories felt too homogeneous and restrictive, so I left them as mixed bags. I wanted the character that fit each sound best for me.
For the locations, I dropped his convention of using the front of a building for the first tone, just inside for the second, any other location inside for the third, and the bathroom for the fourth. This again felt too restrictive. I instead expanded these areas significantly (e.g. -ān became an entire street). I did try to logically connect each set of 4 locations, so that jumping from 1 to 2 to 3 to 4 felt natural. Luckily, with some practice, I found that each location became strongly associated to its ending/tone.
Now that’s out of the way, let’s dive in with a few examples.
1. shòu = thin/slender: Sheldon Cooper is running around the practice soccer field behind my high school. He happens to be a thin guy, so I just exaggerated that feature.
2. jiāng = ginger: Jillian Michaels is bashing a ginger root into the front steps of a nearby house. Apparently it wasn’t doing its workout properly.
3. zhēng = steamed: George Costanza is steaming food inside the hospital lobby.
At first glance, the system may seem daunting. If I tried to use it before I started learning memory techniques, I might’ve given up because it wasn’t producing immediate gains. That said, if there’s one thing mnemonics have taught me, it’s that building effective systems can take time. At first, it may take a while to translate a word into its image+location pair, but it definitely gets easier. And with a few repetitions, those words start to seep into long term memory. That’s the goal, of course. After a point, every word should feel obvious. There should be no need to reference those original images. As soon as I’m able to recall a term automatically, I ditch the mnemonic. Most of the words I’ve encoded have already reached this point. Put another way, these methods are about helping me get words into my brain quickly and then keep them straight initially. They won’t help me fluently whip out old words in conversation—only extensive conversation practice can do that.
I don’t encode everything this way. As usual, the goal is merely to boost learning efficiency, not to impose a rigid structure which may or may not always be most efficient. For instance, for words I still remembered from my one semester of Chinese, I didn’t go back and create new images. I tend not to encode cognates either. Jiānádà (Canada), for example, doesn’t get any images.
This is all well and good, but one issue irked me. Many Chinese terms consist of a pair of words. For example, “to take part in” can be translated as cānjiā. How should I handle these compound words? Nowadays, I usually deal with them in one of two ways.
Method 1: I break down the term into its parts, and memorize each separately. Ideally, they intuitively fit together in a way that’s easy to remember. This is the case for cānjiā, which breaks into cān (to join) and jiā (to add). It’s not too big a stretch to imagine that by joining something and adding to it, you’re taking part in it. Thus, I imagine Aziz Ansari joining a line of plants in my neighbor’s garden. Meanwhile, at the marina, Jillian Michaels is struggling with math addition problems. The upside of this method is that it forces me to learn the component parts, so I’m more likely to perk up when they appear in new contexts. Future words containing cān or jiā, for example, are less likely to slip by unnoticed. The downside is that it takes more time to implement and more time to recall the words initially, since I must jump through two loci. Thankfully, a few repetitions usually resolve this fluency issue.
Method 2: Alternatively, the pair of words may not make intuitive sense, or one of the two may not really be used outside of this one term (so the benefits of encoding it are relatively minimal). Instead of creating separate images, I’ll tack an extra image onto the main one, or—more commonly—I’ll just leave it out and hope it sticks with review). To give an example of the former, yíngyǎng means nourishment or nutritious. Literally, it breaks into “seek + cultivation.” This makes sense, but—as far as my naïve self can tell—the utility of knowing yíng as “seek” is pretty minimal. So, I went through the usual process for yǎng, imagining my Grandmother Edie cultivating crops across the street. Then I added in that she was using a ring (which sounds like ying) to do it.
Method 2 is usually my go-to, because it’s faster and I’m often able to remember non-encoded parts with review. If I find myself forgetting those parts, however, that’s when Method 1 comes back into play.
On a separate note, I chose these palaces before much of the encoding I did in medical school, and many ended up overlapping heavily with my medical palaces. This turned out to not be as big a problem as I’d have guessed (generally, I never “overwrite” palaces; see “Should I reuse palaces to learn new information?” in the FAQ). For instance, despite the fact I used my high school for both -ou and gastrointestinal pathology, I rarely get images confused. My guess is that this works in part because the Chinese words are widely dispersed, whereas the medical images are concentrated. I also visualize them at different angles.
I write these mnemonics into Anki cards to review them. As usual, mnemonics or not, review—ideally via spaced retrieval practice—is essential. I probably sound like a broken record, but when it comes to long-term memory, spaced repetition is the single most important learning concept I’ve encountered. That goes double for language vocabulary. I don’t think that’s emphasized enough in a lot of memory training material. Even if I have fantastic, memorable images, I usually end up forgetting them if I don’t review. Two to three reviews usually gets them to stick long-term.
This all comes with the caveat that I'm not yet fluent in Chinese, so I have no proof this method works in the long run. Like I've said, however, I do think it's a great way to pick up new vocabulary, although of course that's just one piece of the language puzzle. For grammar, I use Integrated Chinese—the textbook set I used in college. Listening and fluency are two hugely important things I still struggle with, although I’m working on it. I haven’t been able to find any Chinese podcasts or radio stations quite at my level, so I do my best to practice speaking with my wife and her relatives. I still need to implement an approach that mimics immersion (i.e. only speaking to my wife in Chinese), which I think will ultimately be necessary if I'm to break out of the intermediate/academic level I've been stuck at for a while. Still a long way to go!