Studying vocabulary (or anything else) with ChatGPT
ChatGPT has made it easy to create customized study materials.
I’ve always loved reference books such as Garner’s Modern English Usage and Brewer’s Dictionary of Phrase and Fable and marveled at the painstaking labor involved in putting such works together. There will always be a place for human-curated reference works because the best of such works reflect their authors’ inimitable expertise and judgment.
In most domains involving the exercise of reasoned judgment, subject-matter experts will always be needed; their creative work cannot generally be performed by an AI system. AI systems are better at breadth than depth. The new large-language models (“LLMs”) such as ChatGPT are astonishingly good at searching for, compiling, and summarizing information, and they will only improve at such tasks. But they are less likely to be able to meaningfully add to the store of human knowledge in domains that traditionally require deep expertise.
If you write academic books or articles and teach in higher-education settings, as I do, you often need to rely on peer-edited publications prepared by scholars with deep expertise in their subjects. For example, since I write and teach about writing, I often need to consult the best reference sources to ensure that my recommendations about word usage are correct. Similarly, as an attorney in the business of giving legal advice and writing litigation briefs, I consult Black’s Law Dictionary rather than Google to check my understanding of unfamiliar legal terms.
For most information-lookup tasks for which people consult reference works, however, superficiality is acceptable. When I encounter a rare word in a book on Kindle, I can highlight the word to see its definition, and that’s good enough for my purpose of learning a new word and understanding the sentence.
That’s one reason why Wikipedia has turned out to be so successful. When the site first became popular, I remember it being spoken of with mistrust and disdain because any amateur can edit its pages. But the site turned out to be accurate enough for most purposes, even if you can’t rely on it as a scholarly source.
ChatGPT has made the process of looking up and compiling information much easier than ever before. If your goal is to educate yourself or create study materials for high school and college students who are still acquiring broad familiarity with a subject, ChatGPT has already superseded all other methods I know for compiling study guides and other learning materials.
Here are a few examples of how I’ve used ChatGPT to create my own study and reference resources for subjects in which I am a relative beginner — decoding Hebrew script (that is, learning to pronounce it without necessarily understanding it), learning rhetorical figures, reading a book about linguistics, and learning unusual vocabulary found in literature (currently, Elizabeth Hardwick’s Sleepless Nights).
Four examples of using ChatGPT as a learning aid
Example 1: Learning literary vocabulary
My query to ChatGPT:
For each of the words below, provide a concise definition of the word as well as two examples of the word being used in literature:
counterpane
forsythia
paratactic
purdah
seraglio
Portion of ChatGPT’s response:
Counterpane: a bedspread or coverlet; Examples from literature:
“She pulled up the counterpane to her chin, and curled her feet inside the sheet, feeling a kind of exhilaration she could not explain, as if she had swallowed a happy pill.” - Chimamanda Ngozi Adichie, “Half of a Yellow Sun”
“I lay down and pulled the counterpane over my face. The last thing I remember was hearing the wind howling outside.” - Emily Bronte, “Wuthering Heights”
Example 2: Learning to decode Hebrew
My query to ChatGPT:
For each letter of the Hebrew alphabet, tell me the most common Hebrew word used in Passover blessings. Include vowel symbols.
Portion of ChatGPT’s response:
Sure, here are the most common Hebrew words used in Passover blessings starting with each letter of the Hebrew alphabet, including vowel symbols:
א - אֲבָהָת יִשְׂרָאֵל (avahat Yisrael) - Father of Israel
ב - בְּרוּכָה הִיא (brucha hi) - blessed is she
ג - גְּאוּלָתֵנוּ (geulatenu) - our redemption
ד - דַּיֵנוּ (dayenu) - it would have been enough for us
ה - הַגָּפֶן (hagafen) - the vine
Example 3: Understanding a technical term
My query to ChatGPT:
What does Walter Nash mean by the term “syntactic recursion” in his book Designs in Prose: A Study of Compositional Problems and Methods?
Portion of ChatGPT’s response:
Syntactic recursion is a concept used in the field of linguistics and cognitive science to describe the ability of human language users to generate complex sentences through embedding clauses within clauses. This property of language is considered to be an important component of our cognitive and linguistic abilities, and it allows us to communicate an infinite variety of ideas and thoughts.
Example 4: Studying rhetorical figures
My query to ChatGPT:
For each of the following rhetorical figures, provide a definition and two examples from literature: Anacoluthon, Anadiplosis, Anaphora, Anastrophe
Portion of ChatGPT’s response:
Anadiplosis: This is a figure of speech in which the last word of a phrase or clause becomes the first word of the next one.
Example 1: “The crime was common, common be the pain” - The Rape of the Lock by Alexander Pope
Example 2: “When I give, I give myself” - Song of Myself by Walt Whitman
As a teacher, researcher, and self-educator, I’m blown away by this technology and hope to see it put to educational uses such as those I’ve demonstrated above. In his recent op-ed about ChatGPT (beautifully titled “Our New Promethean Moment”), Thomas Friedman included a quote that perfectly describes the experience of using ChatGPT for the first time:
To observe an A.I. system — its software, microchips and connectivity — produce that level of originality in multiple languages in just seconds each time, well, the first thing that came to mind was the observation by the science fiction writer Arthur C. Clarke that “any sufficiently advanced technology is indistinguishable from magic.”
Ryan McCarl is a founding partner of Rushing McCarl LLP and the author of Elegant Legal Writing (forthcoming, U. Cal. Press 2024). He is also an adjunct legal writing professor at LMU Loyola School of Law and former research fellow in AI Law and Policy at the UCLA School of Law. Second Stage is his personal blog.
Subscribe for free to receive future Second Stage posts by Ryan McCarl:
Photo by Mahdis Mousavi on Unsplash