Implementing AI in English Language Arts

Note: Examples provided in this blog post use ChatGPT 3.5

Introduction

Odds are, when you think of AI in English Language Arts, things that might spring to mind are limiting students’ critical thinking, receiving AI-written assignment submissions, or stressing over ways to create assignments that aren’t so AI reliant.

However, when used correctly, AI can easily transform the way students learn and how you prepare lessons. It might even help them when teaching them how to write.

This school year, during my work as a reading interventionist, I decided to make use of GenAI (generative artificial intelligence) platforms such as ChatGPT, Google Gemini, and Claude to maximize the level of intervention support my students received. To say that AI saved me time is an understatement.

It drastically helped my students succeed, and it even gave me an opportunity to show the high school English Language Arts teachers how to use it to their advantage.

What follows are tips I provided to the teachers.

Learn the basics of prompt engineering

One thing I stressed to teachers is learning what prompt engineering is and how it works, especially when prompting Claude, for example, to do something in particular.

As discussed by AWS, prompt engineering is the practice of crafting specific text instructions to guide and enable GenAI to produce high-quality, relevant outputs. This process does involve a lot of trial and error; however, we can easily simplify that process.

To summarize IBM’s “Tips for writing foundation model prompts“, a strong prompt includes one or more of the following:

• Instruction
• Context
• Examples
• Cue

Let’s put this into practice by providing an example using ChatGPT (although you are more than welcome to use the GenAI platform of your choice).

Suppose that you are a 10th grade English Language Arts teacher looking to develop sentence stems that your ELs (English Learners) can use for journal writing. In reviewing your information on your ELs, you know that some of them have a basic understanding of the English language, and while they are good at writing about everyday things, they tend to struggle when responding to literature.

You decide that you want to prompt ChatGPT to help you with this task, so you decide to provide ChatGPT with the following prompt:

I’m a 10th grade English Language Arts teacher looking to provide my English Learners with sentence stems they can use for journal writing. My English Learners have a basic understanding of the English language but struggle when responding to literature. Provide some sample sentence stems that might be useful for them.

Now, we prompt ChatGPT with this request, and it provides us with the following:

Now, you might be wondering what would happen if no context was provided in our prompt and we opted for something simpler. Let’s assume that we use the following prompt instead:

Provide some sample sentence stems that might be useful for English Learners responding to literature.

Prompting ChatGPT with this request, we get the following:

See the difference?

It’s important to ensure that while the output might be useful, you should carefully review it to ensure it’s not hallucinating; that is, it’s not making things up.

Practice using AI to analyze literature

Once you’ve gotten the basics of prompt engineering down, the next critical step in implementing AI in English Language Arts is practice, and what better way than to ask AI to analyze literature?

The purpose of doing this is more about finding ways to help students understand difficult or challenging texts, so it helps seeing what GenAI can do to assist you in this endeavor.

Much like our prompt engineering example, we’re going to use ChatGPT to analyze a poem. In this example, I’ll be using Pablo Neruda’s “Tonight I Can Write“.

There are two ways you can do this:

• You can flat out ask ChatGPT to analyze the poem (this might create hallucinations).
• You can copy/paste the poem into ChatGPT and ask it to analyze the poem (this guarantees a little more accuracy).

To Illustrate this, let’s demonstrate how both outputs would look.

This example shows me prompting ChatGPT to analyze the poem:

While the analysis is okay, you’ll notice there’s already an example of a hallucination. For Imagery, it uses the examples of “nights are endless” and “stars watch and tremble”. However, “stars watch and tremble” is nowhere to be found in the poem.

The rest of the analysis, however, seems to be good. What happens, then, if we copy/paste the poem and then prompt ChatGPT to analyze it?

See for yourself:

Similar results, but there doesn’t seem to be any evidence of hallucinations. It’s always important to ensure there are no hallucinations in GenAI’s output; otherwise, you risk lowering the quality of your lesson or activity.

Use AI to help students analyze or understand literature

Once you’ve figured that the analysis is good enough or serviceable, the next step in implementing AI in English Language Arts is using it to help students analyze or understand literature.

Let’s return to our English Learners who have a basic understanding of English. They may struggle with reading poetry, so you might want to use GenAI to create close-reading questions that help them analyze or understand “Tonight I Can Write”.

Let’s ask ChatGPT to do that with the following prompt:

I’m a 10th grade English Language Arts teacher looking to provide my English Learners some close-reading questions for poetry. The poem they are about to read is “Tonight I Can Write” by Pablo Neruda; my English Learners have a basic understanding of English, so they struggle to analyze or understand poetry. Provide me with some close-reading questions they can answer as they’re reading the poem.

Let’s see what ChatGPT outputs:

Note: The response was longer, but I had to make the screenshot smaller to save space.

The questions don’t seem to be too bad and can serve as a starting point if you’re looking to create your own. Of course, questions aren’t the only thing you can ask ChatGPT to create.

Conclusion

In sharing these tips with the English teachers I worked with, they had a better understanding of how not only to implement AI in English Language Arts, but also to leverage its use in creating instructional items tailored to their students’ needs.

Remember, though, that it’s important to check GenAI’s output for accuracy to avoid hallucinations.

How have you implemented AI in English Language Arts? Let me know in the comments!