The next interview: Mike 'Pleco' Love

[roddy]

Following on from the successful interview we did a couple of months ago with Julie of the Huge Oxford Chinese-English Dictionary, I've been roaming around the CSLsphere looking for people to strong-arm into doing similar events. All my dreams and Christmases came true at once when Mike Love, the man behind the much-loved Pleco products agreed to play along.

For those of you know don't know, for the last decade Pleco have been providing us learners of Chinese with Chinese dictionaries (and associated flashcard and reader functions) on various hand-sized devices - from Palm PDAs through to Windows smartphones and all the way up to iPhone and in the future Android and no doubt others.

Mike himself has been posting on these very forums for the best part of five years, both about his own offerings and about areas you'd expect him to know plenty - handheld devices and dictionaries.

So if you've got any questions you'd like to put to Mike on any of those topics, post them in below and once we've got a decent-sized chunk of queries I'll tidy them up and we'll get them answered.

[kdavid]

I've got one question, and then a question/idea.

First, is Pleco available for other non-Apple phones. I have the Nokia X6. Can I download Pleco for that?

Second, I've been using a ZDT/Anki combo for getting new words into my long term memory. I think a good idea for flash card software would be a short-term memory to long-term memory automatic conversion.

When I started using Anki, I wasn't impressed because it didn't help me remember anything. I input words, had trouble remembering them right away, and by the time my next review came around, I had forgotten them all. That's when I realized that Anki is best for keeping information in your long-term memory, and horrible for getting it into your short-term memory.

What I started doing then was pound new material into my memory with ZDT. After I had reviewed a bunch of words several days and felt comfortable with them, I would move them over to Anki and let Anki remind me when to review. This is great as it keeps me from having to review so many new words each day.

My idea for this function would be an all-in-one option where you input new words, pound them into short-term memory, and once there, you can hit "move to SRS deck", which then does what Anki does. It would be nice if Pleco did this as I could then keep all of my learning in one piece of software instead of having to create new files, convert to Anki from ZDT, etc.

Does that make sense? If not, perhaps it's only useful for me. We all learn differently.

[Gleaves]

I'd be curious on his take on the state of Chinese learning today versus 10 years ago when he was just getting Pleco started.

The Chinese learning tools available today are obviously much more powerful than ever before. Have you seen an increase in the quality of Chinese students as a result of new technology? Do you think new technology has helped to "retain" students that might have previously dropped out due to frustration? Do you have any concerns that tools like Pleco might be getting so good (with the advent of effective OCR, etc.) that they could actually become a crutch in studying?

(Disclaimer- I love Pleco. I recently lost my ipod on a trip. I bought a new one the day I returned home just to have Pleco back).

[roddy]

Few from myself . . .

What's it like dealing with the dictionary publishers, particularly those in China? Are they geared up to deal with people who want to license their content for various reasons, or is it a combination of suspicion and failure to understand why you didn't just steal it? How has this changed over the years?

You've quite possibly answered this one already half a dozen times, but how are we getting along with Pleco for Android, the Web, the desktop computer, Java, Multivac and upright arcade cabinets?

If you had input to the dictionary editing process, what would you want to see happening for your purposes - ie, the electronic, for non-native speakers, market?

[gato]

How many programmers are actually involved in the making of Pleco?

How is Mike Love's Chinese nowadays?

[anonymoose]

What are some of the most irritating jokes people make about his name?

[gato]

What are some of the most irritating jokes people make about his name?

Is there going to be a Beach Boys reunion?

[roddy]

Oooh, here's one. What products do you think Pleco will be selling in 2020?

[dreamon]

My question: is there (or planned to be) a C-E dictionary specifically for memorization? This is not the same as a learner's dictionary, nor a translator's dictionary. A learner's dictionary may have perhaps 50000 entries, but lots of explanation and example sentences for each entry; a translator's dictionary has upwards of 200,000 entries, may not have pinyin but has all meanings including literary, technical etc. In both cases, an entry is a unit of explanation. In a memorizer's dictionary, an entry is a unit of memorization, and it must be as small and as catchy as possible. Like in Heisig's book or on a good flashcard. A memorizer's dictionary may still be quite comprehensive, if all units of memorization form a directed acyclic graph that, once traversed, takes the learner all the way to fluency. Such a dictionary should be invoked when a user presses the "create flashcard" button on a vocabulary item, or when a user imports flashcards without a definition. Does this make sense, as part of Pleco?

Edit: One feature that a memorizer's dictionary should have is multiple entries for each word + pinyin combination, i.e. "duplicate" (partial) flashcards. Each partial flashcard covers only one meaning. Of course, the problem is how to pick one of several possible flashcards automatically, when loading new vocabulary. That should probably depend on other flashcards being loaded, as well as frequency (or context, if known).

[abcdefg]

What I started doing then was pound new material into my memory with ZDT. After I had reviewed a bunch of words several days and felt comfortable with them, I would move them over to Anki and let Anki remind me when to review. This is great as it keeps me from having to review so many new words each day.

I use this same method, but don't really mind using separate programs.

[LA Guy]

FYI,

Here is a link to a Pleco / Android vs IPhone discussion ( July 2010 ) on the Pleco forum where Mike Love coyly say's it might be in the works but why not go for an iPhone.

http://www.plecoforums.com/viewtopic.php?f=2&t=2409

I would also like some feedback on anyone using the Android OS to learn Chinese !

LA Guy

[mikelove]

So I guess I'll just answer these here - thanks for the interesting questions, everyone! (surprised there aren't a few more but I guess that's what I get for being so darn accessible :-) )

First, is Pleco available for other non-Apple phones. I have the Nokia X6. Can I download Pleco for that?

Nope, though an Android version should be available in the first half of 2011. We do technically still sell on Palm OS and Windows Mobile as well as on iPhone/iPod/iPad, but they're a negligible portion of our sales (something like 4% to iOS's 96%) and we haven't really been updating them regularly - Palm we officially end-of-lifed a few months ago, WM will probably get one more minor update with some under-the-hood improvements ported over from iOS but is unlikely to see much attention beyond that, so I wouldn't purchase either of those types of device to run Pleco at this point.

Second, I've been using a ZDT/Anki combo for getting new words into my long term memory. I think a good idea for flash card software would be a short-term memory to long-term memory automatic conversion.

When I started using Anki, I wasn't impressed because it didn't help me remember anything. I input words, had trouble remembering them right away, and by the time my next review came around, I had forgotten them all. That's when I realized that Anki is best for keeping information in your long-term memory, and horrible for getting it into your short-term memory.

What I started doing then was pound new material into my memory with ZDT. After I had reviewed a bunch of words several days and felt comfortable with them, I would move them over to Anki and let Anki remind me when to review. This is great as it keeps me from having to review so many new words each day.

My idea for this function would be an all-in-one option where you input new words, pound them into short-term memory, and once there, you can hit "move to SRS deck", which then does what Anki does. It would be nice if Pleco did this as I could then keep all of my learning in one piece of software instead of having to create new files, convert to Anki from ZDT, etc.

Does that make sense? If not, perhaps it's only useful for me. We all learn differently.

This would actually be very doable with our current flashcard system thanks to its support for multiple profiles / multiple categories per card. It'll require a bit of manual-diving to figure out all of the settings, but basically you'd create one category for all of your SRS cards, create one profile for non-SRS and one for SRS, configure the non-SRS profile to exclude cards from the SRS category via the Card Filters / Category Filter option, then configure the SRS profile to require that cards be in the SRS category through the same option. New cards would show up under that non-SRS profile, but when you were ready to sweep them into long-term memory you'd find them via the "Search Cards" command (look for cards with a certain # of correct answers) and add them to the SRS category from search results.

This should all get a good bit easier in the long-awaited Grand Flashcard UI Cleanup that's (hopefully) coming in our next major iPhone update, though it's a few months away...

I'd be curious on his take on the state of Chinese learning today versus 10 years ago when he was just getting Pleco started.

The Chinese learning tools available today are obviously much more powerful than ever before. Have you seen an increase in the quality of Chinese students as a result of new technology? Do you think new technology has helped to "retain" students that might have previously dropped out due to frustration? Do you have any concerns that tools like Pleco might be getting so good (with the advent of effective OCR, etc.) that they could actually become a crutch in studying?

(Disclaimer- I love Pleco. I recently lost my ipod on a trip. I bought a new one the day I returned home just to have Pleco back).

Thanks!

I don't know if I'd say that the "quality" of students has increased, but their willingness to dive into real-world Chinese more quickly and more determinedly certainly has, in no small part due to their ever-increasing numbers and the greater opportunities available for them to do interesting things in China - it's much easier to get yourself in a situation where you're speaking Chinese with real people outside of a sanitized language school enviornment than it was 10 years ago, and much easier to get access to vast amounts of real Chinese-language content (songs, movies, ebooks) outside of China. Even in classroooms there's been much less tedious copying of characters / vocabulary drilling than there used to be and much more interest in learning from books / videos / etc.

And in those situations I think a "crutch" can actually be a very good thing - the sooner you graduate to reading stuff that Chinese people (even Chinese elementary schoolers) actually read, the better, because even if you don't retain all of the vocabulary you see, you're certainly going to retain some of it, and you're going to be getting more and more accustomed to "thinking in Chinese." Ideally you should always be working with Chinese that's just a bit above your current skill level, because that's the only way you're going to really improve. Even outside of Pleco, the "crutch" effect of being able to walk around with a comprehensive Pinyin-English dictionary in your pocket - something that everyone does now but that 10 years ago had roughly the same stigma among 老外 as wearing a calculator watch - can embolden a lot more people to try things like engaging in Chinese conversations with strangers.

The most magical moment in language learning (at least for me) is when you read or hear something and translate it without having to think about it in English. It's not even a one-time thing, you get echoes of it every time your skill level increases a bit and you get access to some new type of content or interaction you couldn't deal with before, but it's not something that comes from cranking through your SRS flashcards every day - you might be able to recall the vocabulary when you're prompted or when you think about it, but the meanings aren't going to start effortlessly popping into your head until you actually get accustomed to using them in real-world contexts.

As far as retaining students, I've gotten quite a few emails from people thanking Pleco for helping them get through their year in China / graduate degree / etc but I don't know how much we really did for them - with or without Pleco you're going to run into walls in the course of learning Chinese, all we can really do is make life a bit more bearable while you're climbing over them. Other companies like ChinesePod and Skritter probably do a better job of making Chinese learning enjoyable - we're more about streamlining the not-so-fun parts. So maybe we've helped a few people stick with Chinese who wouldn't have otherwise, but if you hate language learning or your resolve was weak to begin with I don't know if we can single-handledly turn things around for you.

What's it like dealing with the dictionary publishers, particularly those in China? Are they geared up to deal with people who want to license their content for various reasons, or is it a combination of suspicion and failure to understand why you didn't just steal it? How has this changed over the years?

There are huge variations between publishers in this, actually.

The worst are the big Chinese presses - 商务印书馆 simply doesn't license anything to anyone as far as we can tell, we'd dearly love to get our hands on 现代汉语词典 or 新时代汉英大词典 but it doesn't look like it's going to happen. I think that's a combination of piracy-related frustration - new 现代汉语词典 editions seem like they show up for StarDict before they hit the shelves - and the fact that there simply isn't enough money in it to excite them; Pleco can pay them huge per-copy royalties but our volumes are tiny by Chinese standards, and Chinese software makers can move hundreds of thousands of copies but can't charge more than 10 or 20 RMB (and nowadays mostly just give dictionaries away for free online).

Smaller Chinese publishers can be very friendly and entrepreneurial, though at a certain point they can get to be so small that they aren't really equipped to deal with foreign companies (the paperwork involved for the banking aspects of that alone is monstrously complicated) or are nervous due to a lack of experience in content licensing. Outside of China, the small publishers are just lovely - excited about any opportunity to make more money off of their content, or in some cases even just to have their content seen by a wider audience - while the big ones are usually just as eager to license their stuff but tend to take forever to make decisions about things.

As far as how things have changed over the years, the big non-Chinese publishers have started asking for a lot more money - they don't really see what we do as being all that different than selling books on a Kindle, and so they've gone from expecting us to pay them royalties to expecting us to collect a commission and pay them the rest. The Chinese publishers aren't quite as demanding in that respect (though there's been a lot of grumbling about currency exchange rates lately) but they have gotten a bit less trusting - it's much harder to get a license on a per-copy insead of a flat-annual-fee basis than it used to be, though that may be as much about the transition in China to ad-supported free online dictionaries as it is about not trusting us to report numbers accurately. (having Apple's monthly reports to corroborate us helps a bit on that front, actualy)

You've quite possibly answered this one already half a dozen times, but how are we getting along with Pleco for Android, the Web, the desktop computer, Java, Multivac and upright arcade cabinets?

Android is coming along, though we've been in a bit of a holding pattern on it lately, both because of OCR (which has been a huge success but which we aren't really using to its full potential yet - our system can recognize full-page still images pretty well too) and because Google's dragging their feet about announcing details on Android 2.3 and 3.0. HP/Palm recently announced that they're throwing away their old UI framework on webOS and replacing it with an entirely new one because of tablets, rendering all current webOS applications more-or-less obsolete, and we think there's a significant possibility Google might do the same thing, both because of tablets and because Android's UI in general is starting to look a bit ugly / outdated - if they do, we'd rather throw away the work we've done on the old framework (not that much, most of what we've done has been under-the-hood stuff) and only support the new one since it makes our jobs a whole lot easier.

As far as a web-based version, I'm optimistic that we'll get some web-based functionality integrated into our mobile apps in 2011 - top priority on that is multi-device flashcard sync / backup - but I don't think a full web-based verison of Pleco is a good business move: there's too much other stuff out there, and too much of it free, and our competitors in that space have vast pools of money (nciku's owned by an enormous Korean internet portal e.g.) and can afford to give away a lot of stuff that we can't.

With desktops, we're planning to release a version of our Android tablet software recompiled to run in VirtualBox on desktops and see how that goes. If we eventually release a full native desktop app I think it's more likely we'd do that on Mac than Windows; there's a lot of overlap between OS X and iOS programming, and while there's also a lot of overlap between Windows Mobile and desktop Windows programming, there are much prettier / easier ways of making desktop Windows apps nowadays that weren't available on WM and so a ported WM app would look extremely ugly and outdated.

If you had input to the dictionary editing process, what would you want to see happening for your purposes - ie, the electronic, for non-native speakers, market?

More information would be my biggest request: more words, more example sentences, more detailed usage information, detailed part-of-speech tagging, measure words for every noun; we can source this information from a lot of different places and combine it together, and we do, but getting all of it from the same editorial team in a way that's designed to go together would really be fantastic. That's more an electronic thing than a non-native speaker thing, but it's an ongoing source of frustration for me - why are we still designing dictionaries around print when most people are now getting them electronically?

I'd love to see a really great character dictionary for foreigners, in fact that's one I'm thinking about commissioning at Pleco since the few titles out there now are uninspiring and most of them don't seem to be licenseable anyway - something that covered both scholarly character etymology info and useful purely component-meaning-based mnemonics. A Chinese-English dictionary of grammar (sentence structures, patterns, etc, organized not as lesson-by-lesson grammar tutorial but as a searchable reference) and one of synonyms / antonyms would also be lovely thing to have - the latter could probably be automatically generated from a Chinese-Chinese one and a Chinese-English dictionary, actually.

Oh, and while I'm making a lexicographical Christmas list, a good frequency list of spoken Chinese word frequency would be great - John Pasden from ChinesePod listed this as one of his "Two Wishes for Chinese Language Instruction" and I agree, if we don't know what words people are actually using then we don't know what words to teach or what words to include in our dictionaries; new words crop up so often that you really almost need a permanent / full-time project to keep on top of them (maybe from the Confucius Institute?).

How many programmers are actually involved in the making of Pleco?

Difficult to say: lots of pieces are contracted out to other individuals or companies - there are probably at least two dozen people's code involved somewhere - but I'm the guy who designs and assembles it all (not always successfully).

How is Mike Love's Chinese nowadays?

Rusty, sadly - I read extremely well, but I hardly ever have natural opportunities for conversation practice and I haven't had the time to set up a regular meeting with a language partner / sign up with a VoIP tutoring service / etc.

What are some of the most irritating jokes people make about his name?

Is there going to be a Beach Boys reunion?

Very few Beach Boys jokes, actually; not many people even know that reference these days, I've made the joke myself once or twice and been met with only silence. Maybe if I were named Brian Wilson...

Not that many jokes in general, in fact, more just general laughs / "interesting name" comments - the frequency of those has a lot to do with where the speaker is from, in parts of the US it's an exceedingly common last name (North Carolina, e.g.) and in other parts it's quite rare, though I can't say as I've ever personally encountered another Love who wasn't related to me.

Oooh, here's one. What products do you think Pleco will be selling in 2020?

Voice recognition, hopefully - speaker-independent Chinese voice recognition isn't quite there yet, and I'm not that impressed with current-generation technology for correcting foreigners' Chinese pronunciation either (saw some stuff the US military had developed for use with Arabic in Iraq and even that was pretty poor), but it's bound to get there eventually. It would be incredibly useful to be able to talk to somebody and then get an instant transcript of what they'd just said which you could read over / tap on words to translate / etc, and that may not be that far away; it'll probably be Pinyin-only at first (or woefully inaccurate in its character transcriptions) but that'll still be enough to work out from context what they said.

I wish I could say machine translation, but I think that's going to take a lot longer to really make practically useful - it hasn't really advanced much in the last decade, statistical approaches like those used by Google / Microsoft seem to do a little better than the older "smarter" ones but they can only go so far.

Of course this is all assuming Pleco is still around in 2020 - we could easily get bought out at some point, it's looking more and more likely that we're going to take the plunge into seeking VC money within the next year or so, both due to the aforementioned increasingly expensive dictionary licenses and because with the smartphone software market booming there's never going to be a better time. Or we could be driven out of business by the proliferation of free apps - it's taking them longer than Wikipedia, but certainly by 2020 I expect CC-CEDICT will beat out any published C-E dictionary both in terms of accuracy and extra content (they're going to get example sentences / parts of speech / usage notes / etc in there sooner or later), and it could be as tough to make money selling Chinese dictionaries then as it is to make money selling encyclopedias now.

My question: is there (or planned to be) a C-E dictionary specifically for memorization? This is not the same as a learner's dictionary, nor a translator's dictionary. A learner's dictionary may have perhaps 50000 entries, but lots of explanation and example sentences for each entry; a translator's dictionary has upwards of 200,000 entries, may not have pinyin but has all meanings including literary, technical etc. In both cases, an entry is a unit of explanation. In a memorizer's dictionary, an entry is a unit of memorization, and it must be as small and as catchy as possible. Like in Heisig's book or on a good flashcard. A memorizer's dictionary may still be quite comprehensive, if all units of memorization form a directed acyclic graph that, once traversed, takes the learner all the way to fluency. Such a dictionary should be invoked when a user presses the "create flashcard" button on a vocabulary item, or when a user imports flashcards without a definition. Does this make sense, as part of Pleco?

Edit: One feature that a memorizer's dictionary should have is multiple entries for each word + pinyin combination, i.e. "duplicate" (partial) flashcards. Each partial flashcard covers only one meaning. Of course, the problem is how to pick one of several possible flashcards automatically, when loading new vocabulary. That should probably depend on other flashcards being loaded, as well as frequency (or context, if known).

It's an interesting idea, but I think it might be confusing to call it a "dictionary" since it's not designed to be used for looking up words; in fact, by splitting up the meanings of words into multiple entries it would somewhat ill-suited for that. This would really almost be more of a custom-designed flashcard deck, one that was frequency-ordered and had a separate entry for each sense of a word.

That could be a lovely thing to have, but compiling it would be a lot of work, since you're basically laying out an entire Chinese course; I'm also worried it wouldn't work that well with the + button, since as you point out it's not clear when you tap on that which sense of a word you want to create a flashcard from.

But I have felt for a while now like we need to add some sort of built-in, official flashcard list set, and with the reaction to the new HSK being so-so at best it doesn't seem like that's necessarily the best thing to base it on, so it might be worth investigating further - thanks for the suggestion.

Here is a link to a Pleco / Android vs IPhone discussion ( July 2010 ) on the Pleco forum where Mike Love coyly say's it might be in the works but why not go for an iPhone.

Oh it's been made considerably more official since then, but I'd still get an iPhone - it'll be a while before our Android version is actually done, and since it's so tricky to optimize around Android hardware (cool performance improvement works great on one phone but is broken on another) some aspects of it may be considerably less smooth / pretty than their iPhone equivalents. But the fact that we have our own proprietary Chinese-dictionary-optimized database engine written in native code should let us run circles around any other Chinese dictionary on Android at least.

Thanks again for the questions, everyone!

[dreamon]

May I add one more suggestion: audio flashcards for the car. Could be based on the "regular" flashcard decks.

[Gleaves]

Thanks, Mike. Thoughtful answers.

Per Dreamon's suggestion, hands-free flashcards by way of voice commands could be very cool.

[dreamon]

Gleaves> hands-free flashcards by way of voice commands could be very cool.

It is fine if they are "feedback-less" or if there is a remote with just two buttons, "correct" and "wrong" (or even one button, "correct"). But the system will have to pronounce the English translations. Which, in the 1st version, could be my own voice records (?).

EDIT: A better set of two buttons is "Correct" and "Pause/Resume".

BTW, I'm just curious: did anyone try to develop a projector on the car's windshield so that the driver can see, say, the map while watching the road? Imagine characters projected this way...

[Deb Thomas]

I have been looking for a robust Mandarin English dictionary for my iPad and have bought the Collins app which is OK but has a few limitations. Does Pleco have any dictionary apps for the iPad that anyone would recommend?

[Daan]

I'm looking forward to 2011 even more now! Thanks, Mike.

[Don_Horhe]

This might have been asked before, but here it goes...

Is there a flashcard deck database that users who own the software can upload to and download from, like Anki's? If not, are you considering developing something similar?

Is there a way to import decks into Pleco from software like Anki? I plan on getting Pleco once I receive my next pay check, and it would be really neat if I could transfer my existing Anki decks into Pleco, instead of having to create them manually from scratch.

[mikelove]

Per Dreamon's suggestion, hands-free flashcards by way of voice commands could be very cool.

Thanks! Voice commands I think we'd actually be more likely to experiment with on Android first, since it has some built-in voice recognition support and iOS doesn't.

It is fine if they are "feedback-less" or if there is a remote with just two buttons, "correct" and "wrong" (or even one button, "correct"). But the system will have to pronounce the English translations. Which, in the 1st version, could be my own voice records (?).

EDIT: A better set of two buttons is "Correct" and "Pause/Resume".

That could also work - we might even be able to generate the English translations with a free online system (though we'd have to either download the audio in advance or require that you have uninterrupted internet in your car).

I have been looking for a robust Mandarin English dictionary for my iPad and have bought the Collins app which is OK but has a few limitations. Does Pleco have any dictionary apps for the iPad that anyone would recommend?

Our free basic app works beautifully on iPad, so I'd recommend that you take a look at that and see if it might meet your needs.

I'm looking forward to 2011 even more now! Thanks, Mike.

Thanks!

Is there a flashcard deck database that users who own the software can upload to and download from, like Anki's? If not, are you considering developing something similar?

Not seamlessly integrated yet, but we do have a Flashcard Exchange discussion forum on PlecoForums.com that serves the same purpose.

Is there a way to import decks into Pleco from software like Anki? I plan on getting Pleco once I receive my next pay check, and it would be really neat if I could transfer my existing Anki decks into Pleco, instead of having to create them manually from scratch.

A number of people have reported success with that, though you might have to do a bit of reformatting of the file you get from Anki before it'll import into Pleco seamlessly - the tricky thing is that not all Anki Chinese flashcard decks use the exact same format. I should really write up a guide / tutorial on that...

[Don_Horhe]

Thanks, Mike!