As my Legal Writing students reminded me the other day, it’s one thing to know a word when reading. It’s another to figure out how to use it correctly in writing.
In this case, the troublesome legal English word is “precedent.” Below are a few sample sentences from my students’ writing with attempts to use the word:
- “According to the precedent case, the police officer’s supervisor issued a ticket to a person who threw a candy wrapper on the ground.”
- “The precedent shows that coffee poured on the ground is not litter while a candy wrapper is litter.”
- “The precedent is the police officer’s supervisor has issued a ticket to a person who threw a candy wrapper on the ground.”
- “One precedent is that a person who threw a candy wrapper on the ground was issued a littering ticket.”
Grammatically, these sentences are fine. Yet, as a native English speaker and teacher of legal writing, the use of “precedent” sounds decidedly off. But why? What you might find yourself saying is, “It just doesn’t sound right.” Yet that feels like an insufficient explanation given that the fundamental ability a non-native English speaker lacks is the ability to know what sounds right.
So what is a legal writing professor of LLM students to do?
I want to suggest that “It doesn’t sound right” can in fact be an appropriate answer, as long as you can provide an explanation as to why. And a great tool for helping you identify that explanation is the iWeb corpus created by the corpus linguists at BYU. Unveiled in May 2018, the 14 billion word iWeb corpus was created by the same BYU people as an improvement on the 560 million word Corpus of Contemporary American English (COCA), which had been the most popular and well-known freely available English corpus to date.
From my perspective as a teacher of legal English to LLM students, however, what makes the iWeb corpus such a helpful tool is not the additional 13.44 billion words. But rather the features and layout the display when you look up a word. It gives you a suite of well-organized, useful information all on one page.
For example, if you type in the word “precedent,” it shows you the following:
- Related topics
- Collocations, organized by Nouns, Verbs, Adjectives and Adverbs
- Links to Youglish.com, PlayPhrase and Yarn, sites that show you video clips culled from YouTube and other sources of people actually using the word in speech
- Clusters: Words that frequently appear before and after the target word
- Concordance sentences: Actual sentences that are aligned so that the target word from each sentence forms a neat column.
By the time you look at all of this, you get a sense of how the word is and isn’t used.
Now, checking back with the student sentences and looking at the Clusters section of “precedent” on iWeb corpus, we see that the word “precedent”:
- is often preceded or followed by phrases involving the verbs “set” or “establish” or by the phrase “there is a” or “there is no”;
- is often followed by prepositions such as “for” and “in”;
- is preceded frequently by adjectives such as “dangerous”, “legal”, “without”, “bad”, and “historical”
Additionally, a search of the phrases used by the students shows no sentences for the following strings:
- “According to the precedent case,…”
- “The precedent shows that….”
- “One precedent is that ….”
And a search for the other string (“The precedent is….”) shows primarily a very different use pattern than the one employed by the student. Instead, “is” is often either paired with an adjective (e.g., “the precedent is not good”, “the precedent is binding”) or with a past participle to make the passive voice (e.g., “the precedent is set”, “the precedent is considered to be”).
Additionally, a search for the string “the precedent is that” brings up a total of 10 results:
Upon examination of the 10 results, it seems that the phrase “the precedent is that” tends be followed by a general statement of principle, i.e., use of general terms and present tense. Contrast that with the student sentences which all refer to the specific facts of the case of the coffee or candy wrapper, rather than a general statement of the principle, and also use the past tense.
In this way, the iWeb corpus can serve as a decent proxy for that intuitive sense of what sounds right that LLM students will often lack. Additionally, if you are able to teach your students to use the corpus, then you can set a new “precedent” for strategies to use when not sure of how to use a word.
My own recent strategy has been to keep a browser tab with the iWeb corpus up on my screen during class. Whenever a word comes up in discussion that makes me think, “That doesn’t sound right,” then that’s my signal to pull up iWeb and plug it in. The secondary benefit is that the more I model this in class, the more the students will think to turn to it when working on writing tasks.
If you have other corpus-based strategies you’ve tried in the legal English or similar context, or other legal English words or phrases you’ve plugged into the iWeb corpus in your work with students, it would be great to hear about your experience in the comments section below.