1, What are academic papers in the humanities and social sciences?

First, we need to distinguish between several common types of articles: essays, commentaries, and academic papers.

Essays typically emphasize personal expression, persuasive language, and narrative ability. For example, a topic like “Urban Changes Through My Eyes” is more like an essay. It can be written with great emotion and vivid detail, but it doesn't necessarily require rigorous data support or academic argumentation.

Commentaries, on the other hand, emphasize stance and realistic judgment. For example, “Urban renewal should not sacrifice the interests of original residents” is a commentary. Commentaries can have a clear stance and incorporate real-world examples, but they usually don't need a complete research framework like academic papers.

Academic papers are different. An academic paper must have a core question and a core argument. For example, if we're writing about gentrification in Harlem, we can't just introduce its history, nor can we simply say  “gentrification is bad”. A better approach to the paper would be: Harlem's gentrification is not just a matter of rising housing prices, but rather a process in which local culture is repackaged by capital, resulting in spatial exclusion and social stratification.

The first step in writing a thesis is not to immediately begin writing the main body, but to learn how to transform the topic into a research question.

A TOPIC is a subject of discussion. A RESEARCH QUESTION is a problem to be studied. They may seem similar, but they are different.

For example, “AI education” is just one topic. It's too broad. You can write about almost anything, but it's also easy to write about nothing clearly. If we change it to whether AI detection tools will have an unfair impact on non-native language students, then it becomes a research question.

For example, “US banknotes” is just one topic. If we change it to How the design of US banknotes reflects changes in national identity and economic order, it gains a clear research direction.

So, what makes a good research question?

First, it must be specific. The question shouldn't be too broad. For example, the topic     “Why did China's democratization fail? ” is too broad and difficult to explain in a high school paper. However, a better question would be  “Why did the collective leadership system during Hu Jintao's era reveal structural fragility during power transitions?” This makes the scope clearer and easier to analyze.

Second, it must be open to debate. A good research question cannot simply state a fact. For example, “What year did the American Civil War begin?” is not a research question because it has a standard answer. But “How did the media shape the narrative of national unity after the American Civil War? ” can be a research question because it requires explanation and argumentation.

Third, it must be supported by available data. No matter how interesting a topic is, it's difficult to write a paper if reliable data is unavailable. Academic writing isn't based on intuition but on evidence.

Fourth, it must be meaningful. This meaning doesn't necessarily mean changing the world, but it should at least explain why this issue is worth discussing. Can it help us re-understand a historical phenomenon, a social problem, an economic mechanism, or a cultural phenomenon?

After identifying the research question, the next step is to make the thesis statement, which is the central argument.

We can use a simple formula to help everyone write their thesis:

Although X seems to suggest A, this paper argues that B because C.

Let's take an example: AI in education.

This is a very topic and it is too big. We can narrow it down to AI detection tools in high schools. Then, we can further formulate a research question: "Do AI detection tools unfairly affect non-native English students?"

Next, we can develop a preliminary thesis: AI detection tools may create academic unfairness for non-native English students because they can confuse unconventional language patterns with machine-generated writing.

In this way, a qualified thesis is revealed.

In the next section, we'll discuss how to find, evaluate, and use evidence.