CAPL, an open multilingual photo dictionary

capl.jpg As a language teacher, I am always making my own materials. Good language teachers do this habitually. We don't only rely on content provided to us; we constantly tinker with materials, re-form them, borrow, adapt, and use anything we can get our hands on to make the language learning process more relevant and interesting to our students. We are essentially "makers" of curricular materials. Images do improve vocabulary acquisition and are essential for the instruction of culture. Simply put, access to good media is limited for teachers. Sure, we all have MS Clipart, but how do you explain a ticket cancelling machine in the unit on public transit with clipart? Or how do you convey what a Döner is? An image is a good place to start. Simply taking images from search engines doesn't always work. There can be too many and finding the right image is often hard when you want to convey a specific cultural idea in a different language. In 2003 I approached a colleague of mine at W&J College about a project idea I had.
Jason Parkhill, now director of academic IT services at Colby College, worked with me to help design, program, and provide content for what we consider to be the first "authentic" picture dictionary. It is a classic academic project that was under funded, too complex for two people, and for which we were unsure of the target audience. We went with the non-sexy name of a "pictorial lexicon" as we thought it best described our efforts to visually represent a set of words in a given language. The images in CAPL are not generic representations of objects, but are photos taken in an unaltered context. Since German is my field, I figured we would start there as a way to attack the problem I had finding high quality images for my German classes. In designing the database, we went with the strategy of providing a limited amount of meta-data to accompany each image. We basically have two categories for each image, an English description, and a short vernacular description. In a world of too much information, limited data can be a blessing. • Try finding an image in our database. It is fairly easy. •Search Google images, our limited meta-data has our images appear all the time when you search, especially when filtered by license. •Browse a category if you don't know the specific word. If you are looking for an image of a specific object in a culture, my hope is that you can easily find it there. Clearly there are other sources for images, Corbis, Getty, and Wikipedia for example, but the amount of information there can be something through which you must wade with knee high boots. Besides Wikipedia, the cost of other photo databases is prohibitive. The German database was well received by the teaching community, logging 6 million page views in a few years. Yes, BoingBoing gets 10 million a month, but for a home grown teacher resource and almost no funding, we feel like we came a long way to make our point about visual aspects of different languages. Last year, however, we got a grant to start to expand the database and re-program it to handle multiple languages. Two very sympathetic IT guys at the college, Jason Pergola and Brad Kita, came on board and put in a lot of time to adapt the database to handle multiple languages and keep the interface simple. We also added some new content and the new database works well. You might even find an image of a Russian banana if you desire to look at one. Once we have some content and an editor, we can expand into any language and are editing new databases currently. Behind the CAPL database has always been the spirit of sharing content so that other educators may "make their own" lessons. We use a Creative Commons 3.0 US License to allow for free non-commercial use and adaption. We actually encourage use and adaption of the images there. Schools, universities, or any nonprofit entity or individual may use these authentic photos for a variety of purposes. Use them on a school web site; put them in a blog, print them on a credit card for all we care. The images come in three sizes from thumbnail to very large and we are constantly expanding. My college (Washington & Jefferson) has long been a supporter of this open content project. Connexions is a great resource database cataloging numerous open content projects for making your own educational materials. My long term hope is that at some point there will be a truly authentic visual dictionary that is multi-lingual and authentic. Current search engines don't have an elegant way to sift through visual content with a cultural filter. Perhaps it could be done with geo-tagging in combination with meta-data, but for now our project does it by hand. We are trudging along with our group of volunteer experts who edit the images we have in our database. It is a type of "slow media" project. As you can see in our database list, there are some starter projects just getting off the ground with less than 1000 unique entries, and we have other languages that have many more images in the database. For now, if you need a CC licensed photo of a particular object, try CAPL. I recently was at the American Council on the Teaching of Foreign Languages (ACTFL) Conference 19 talking about the pedagogical reasons for using authentic images over generic images and how teachers can use CAPL to make their own multi-media learning materials. I have provided some examples of how teachers are already using these images in the traditional language learning classroom, but I would love to hear from some creative readers what they think they could do with the images in an electronic learning environment. How would you make a web based vocabulary learning game out of the images? Would you adapt it for a mobile device? If so, which platform? What are your experiences with educational media? (From film strips to pod casts) I am exploring options for mobile learning using authentic images now and will start once we have our funding in place. In the academic world, we might say "There's a grant for that." Knock, knock, Gates foundation.


  1. I think this is a great idea. I also think, though, that there’s a huge opportunity here that isn’t being fully exploited.

    This type of database is crying out to be wiki-like, so that people can contribute entries, help with translations to other languages, add metadata (keeping it simple is great, but some entries I viewed lacked any context), etc. This needs to be possible to do rather casually – like on wikipedia, there can be your group of core volunteers, but it should be dead simple for others to contribute right on the site.

    Case in point – I would love to contribute to a Thai section, with my own photos from Thailand that I would CC license for a good cause. I probably would only do a handful to start, but just having those there might encourage others to contribute, if it’s easy to contribute.

    Regarding metadata… one problem I often encounter when trying to translate things to or from Thai is that the various dictionaries I use (online and off) are written with different English dialects in mind. British, American, Canadian, and Australian words for the same things can vary wildly. Even within one region, there can be lots of different words for the same thing, different ways to express an idea, and so on. Very few translation dictionaries address this problem, including this.

    This has much wider potential use than as a resource for language teachers – it would be great if it were accessible for self-learners as well. Right now, it’s completely one-sided – it will be most useful for English speaking teachers.

    A final thought… the navigation now is pretty clunky. The animated menu effect is cute, but not very friendly to use.

    1. When we started this project that was the main question. Do we created a system that is edited or a wiki style database where everyone can contribute.

      I agree we could have many more images if we opened it up but I think we would have a whole host of new problems in doing that including monitoring spam photos. Perhaps Wikimedia will create a visual lexicon section that is multi-lingual some day?

      Our approach has been to keep it simple, with limited search data and that model has worked in the way we wanted. It is one-sided as you indicate as the English descriptors are in American style English. That does limit it a little I admit, but we see the photos as a resource for teachers mostly to use in their classes.

      I don’t think any of the images will teach the language by themselves. They need a pedagogical context. And yes, some of the images are sloppy @dr. It is a work in progress and is growing as I type (as evidence by some of the fledgling sections)

  2. I only looked up one thing – coffee – and find that the entry for “coffee grounds” is a closed bag of something that *might* be coffee grounds, but also might be whole beans, or oatmeal, or nuts and bolts. Another photo, “ground coffee”, is also an opaque container, in this case a plastic jar. This seems excessively inaccurate (even sloppy) for something that is supposed to be helping language learners.

    1. I had the same negative experience for the only thing I looked up, “oatmeal”. Even if this is not fully open and wiki-like, there should be an easy way for users to help correct things, flag things as inaccurate/extremely unclear, something…

  3. You may want to add the possibility of collapsing all regional variations of a language into a single language entry.

    Try searching for “casa” (house) in ES-MX. Nothing. Search for the same term in Spanish from the Southern Cone and Central American and you get different results, most of which are valid for all Spanish variations.

    1. Thanks, good idea. Spanish is an area where we are experimenting due to the linguistic and visual diversity of the Spanish speaking world. Right now we have projects for Peninsular, Mexican, Caribbean, Central American, Andean, and Southern Cone varieties underway.

      1. What’s great about bb and a few other blogs is that in the mix of readers you’ll find some fantastic photographers, many of whom I’m sure would love to take a simple, beautiful, high-resolution image that effectively represents the object.

  4. The French area clearly needs a Québec section, since you will get a lot of submissions from Canada, where they really don’t speak a standard metropolitan french (whithout admitting it !).
    For example, An entry for the word “gosse” would be totally misleading : it means “kid” in standard french, but “testicle” in Québec french ! Same thing for knickers, named “culotte” in french and “bobette” in Québec french.

  5. I was a student when you first began this project, so it is interesting to learn about the progress that you’ve made, congratulations!

    As far as gaming or adapting to a mobile device, it would be cool if you could do something to align with the current trend in location updating. I was recently in China and used FourSquare to “check-in” to a variety of locations, ranging from technology parks to the Great Wall. As a starting point, you could even design a few badges that create awareness for CAPL. Later on, it could be an app where folks can take a picture of something and you can name what it is (can photo recognition software help with this? I’m not a techie, literally I don’t know). Or, integrate the images with some sort of country/region specific scavenger hunt.

  6. How to explain a “Döner”? Get them drunk first, then feed it to them. No words necessary (and now I’m craving a doner kebab, and I’m not even drunk).

    I do like the “KISS” approach to the lexicon, though…

  7. “Ticket cancelling machine” is a rather rough translation. Its a machine that stamps the time and date on a ticket or punches holes in it. You could call it a mechanical conductor.

  8. “Ticket cancelling machine” is a rather rough translation. Its a machine that stamps the time and date on a ticket or punches holes in it. You could call it a mechanical conductor.

    1. “You could call it a mechanical conductor.”

      Except that ticket canceling machines don’t give directions, sell tickets, or announce the next stop. So they are not really mechanical conductors, except in a very limited sense.

  9. This is a great idea. I love it, but I have a suggestion that would improve the experience – there should be a “browse all” selection for each language in the Browse section. At the moment, it’s a bit of a hassle to go through all of the images at once in a given language.

    As for electronic media, I use podcasts in my own language classes fairly regularly. Mainly I use news reports (both audio and video) to give examples of authentic speech, native-level speed and vocabulary usage, and current events. Podcasts are also good for demonstrating the variety of English accents in a natural context. ELLLO has a huge amount of material for English listening practice, but it’s not very high-level stuff.

    Good luck with your project!

Comments are closed.