The German phrase “Ich verstehe nur Bahnhof” means “I understand only train station.” To a native English speaker, this might sound like nonsense. To someone who knows German, it means that the speaker doesn’t comprehend something; it’s the equivalent of saying the English idiom “It’s all Greek to me!” Even English speakers, though, might not know the words garrulous (chatty), inchoate (rudimentary) or cognizance (knowledge).
So how do scholars who study language figure out which words you really need to know?
This is exactly the question that Jamie Rankin asked when he began his career in second language learning and vocabulary. It also served as the thesis for his lecture “How can I learn all these words? Research-based strategies for L2 teaching and curriculum development.”
“I set out to answer some questions,” said Rankin, who visited Binghamton for his lecture on Sept. 27. “How many words do people know? How do those words relate to textual coverage? How many words do people know and how many of those words cover a given amount of text? What’s the most effective way to help students learn these words?”
A lot has changed in the last 50 years of teaching language, he explained. Early on, there was a focus on grammar, mainly in the hope of translating antique materials back into English. Then, in the 1960s, some classrooms focused on habit formation by having students mimic their teacher’s phrases, a practice that became known as “audiolingualism.”
More recently, professors have focused on preparing students for cultural immersion.
“If you’re doing grammar translation, if you’re doing audiolingualism, all you’re doing is learning rules. You’re not really learning how to speak,” Rankin said. “Then you have communicative language teaching, which sort of grew out of that, where the whole point was to use the language for spontaneous communication among people. And I would say that for most colleges and universities, this is the way languages are taught. The point is to be able to communicate competently in the language.”
This method of teaching focused on ensuring linguistic proficiency in more than one arena — the ability to not only say something grammatically correct, but socially correct, in a way that flows cohesively. It even aims to improve your aptitude to navigate conversations in which you may not have the language needed to answer as you normally would.
But there’s a problem Rankin continued to run into, even with this technique: the sheer number of words that can be taught and the process of prioritizing some over others. To focus on these cultural aspects of learning, many textbooks neglect vocabulary — and this is true for every language.
“Now that doesn’t mean that the authors just wanted you to learn [random] words and they were trying to be cruel,” Rankin said. “They were choosing vocabulary with the intention of making it useful for a cultural topic, not because it was what’s actually going to be found in a text.”
The science of language
In the Oxford English Dictionary, there are about 600,000 words. Many are archaic or based on regional dialects. How do teachers and students even know where to start? To answer that question, the art of language turns to science.
First, researchers compile a “corpus,” a collection of sample texts that give a user an idea of how a language is actually used. Instead of standard academic items, the idea is to include more authentic sources, such as articles from newspapers, emails or tweets, covering a representative range of materials. At one point, this was done by hand, but computers have sped up the process.
Then, researchers take the words and split them into word families, or groups of words that are variations of each other — for example, “agree,” “agreement” and “disagreement” — and count how often they appear. Finally, they find the “lemma,” which is the root version. This final set of words is compiled into a list.
One such corpus, the Brown University Standard Corpus of Present-Day American English, compiled the words in a collection of 500 total texts. By evaluating exactly how frequently the most common words appear in those texts, researchers were able to determine that the top 1,000 words would cover nearly three-quarters of the text.
“The first thousand words that were determined to be the most frequently-used words in that corpus covered 72% of those texts,” Rankin said. “I thought ‘This can’t be right.’ For one thing, in the top 1,000 or so, there is not one article of clothing. In every first-year textbook, you get a stick figure with 30 different articles of clothing. You didn’t have the words that you expected from commercial textbooks, but you did have words that no commercial textbook has during the first year. You have lots of words that I didn’t really think about learning until 10 years after I started learning German.”
Many of these words, it turned out, were glazed over in classic classroom structures. Words like “the,” which appears at the very top of the list, or prepositions such as “out,” “if” and “about” were all included in the top 1,000, along with a number of other unexpected parts of speech.
An interactive curriculum
The revelation that the language model being taught in beginner-level classes didn’t match up to the science led Rankin to a project. He bought 1,200 color-coordinated Post-it notes and set to work organizing a curriculum that focused on this set of high-frequency words. Carl Gelderloos, associate professor of German Studies in the Department of German and Russian Studies and the moderator of the lecture, has used the finished project in his classes. Rankin has even developed a special version for Binghamton students.
The completely online, interactive curriculum, der|die|das, is for beginner-level German students. The goal, says Gelderloos, is to ensure that students leaving the classroom, even those with no prior experience, would be able to completely navigate both conversation and cultural barriers.
“By the second week of German, students are able to have conversations with each other, talk about their interests. By the end of two semesters, they’re really able to handle most situations, and even think about studying abroad in German at a German-speaking University,” he said.
The “textbook” is now used in universities across the world. The biggest difference between der|die|das and other commercial works is its organization. Instead of ‘semantic’ lists of active vocabulary (words that are similar in meaning, like articles of clothing), the program used thematic sets, which are drawn from the different elements of language — verbs, nouns and adjectives — in order to tell a story and engage with the topical focus of each chapter.
The program also structures the coursework with regular classroom reinforcement and various types of learning, from visuals to sound and sometimes even touch. Although it focuses on the top 1,000 words of the corpus, there is supplementary material that leaves students even more prepared.
Rankin continues to work on the project, refining the content to meet the needs of the language. He hopes to release a second version that continues to teach the higher-frequency words that comprise the next set of words in the corpus.
Meanwhile, however, Gelderloos will continue to teach using the program, which has demonstrated effectiveness, and hopes that students seeing Rankin’s research will be reaffirmed in the importance of understanding language on a deeper level.
“Teaching language is probably more art than science, but teaching it well really depends on being in dialogue with the latest research,” Gelderloos said. “I think it will help students see that there’s a lot of work that goes on behind the scenes in terms of teaching language. Languages are living things. And if you can actually gain proficiency — it’s a cliché — but it opens up worlds to you. For the students that I work with, I know that it’s going to be a really exciting idea.”