They Say I Say

Getting past the opening argument trap

Most students write their introduction as a blank room, then walk in and claim the whole thing is theirs. That never lands. Readers want to know where your idea sits among the people already talking about it. The They Say I Say framework exists because that transition is the hardest part of academic writing, and most writers blow it by just stating their own position without any setup. I have spent years grading papers where the writer's thesis is fine but unreadable because they never showed the conversation they're entering. You can fix this by learning the actual moves, not the marketing version of them. The core idea is simple enough that you could explain it in a sentence: state what someone else thinks, then state your response. But the actual execution is where things go wrong almost immediately.

They Say I Say: the mechanics nobody talks about

The framework has four real components that work together, and most people only use the first one. You need to establish the "they say" portion fairly, which means representing the opposing or source position accurately enough that its actual author would nod along. Then you add the "I say" portion, which carries your argument. Between those two is the pivot, usually a contrast word like but, however, or yet, followed by the reason your response matters. The template Graff and Birkenstein built around this includes specific moves like opening with a quote and framing it, summarizing a position in your own words before responding, acknowledging where a source might be right before pushing back, and using consensus statements as springboards. Each move handles a different relationship between your voice and existing scholarship. I find the one most people skip is the move where you grant a point to the other side before pivoting. That concession move actually makes your argument stronger because it signals intellectual honesty instead of strawmanning. Here is where I hit a wall recently. I was helping someone edit a literature review for a sociology project and we kept running into the same problem: the writer had five sources but no actual conversation between them. Each source got its own paragraph, listed like a grocery receipt. Nobody was responding to anyone. I had her rewrite the opening of each paragraph to reference the previous source directly, using phrases like "While Chen argues that structural factors dominate, Davis pushes back by pointing to cultural variables." That single change turned a disconnected list into an argument. It took about twenty minutes to fix a section that had been giving us grief for three weeks.

Another thing that always surprises beginners is how much the quality of your "they say" portion determines whether your "I say" gets taken seriously. If your summary of someone else's position is weak, vague, or slightly inaccurate, readers will either stop trusting you or start fixing your mistakes in their head instead of following your argument. I once watched a grad student lose an entire chapter's credibility because she summarized a methodology she clearly hadn't read closely. She wrote "Smith et al. conducted a longitudinal study" when it was actually a cross-sectional design. The reviewer caught it in thirty seconds and questioned everything after that too. The templates in the original book are useful starting points but they become transparent if you use them verbatim. Phrases like "Many people believe that X" or "In response, I argue that Y" get flagged by anyone who reads enough academic writing. The better approach is to internalize the moves and then write them in your own discipline's voice. Engineering papers summarize prior work differently than history papers do. Your framing should match the expectations of the field you're writing into. There is a real downside to relying on this framework too heavily, and it is worth knowing about before you commit to it. The pattern can make your writing sound mechanical if you lean on the same contrast structures repeatedly. When every paragraph follows the same "they say but I say" rhythm, readers tune out. I have seen students produce perfectly structured papers that read like form letters. The solution is to vary the order sometimes, or to embed your position more subtly within the source discussion rather than flagging it with explicit pivot words every single time.

Another limitation is that the framework works best when you are responding to actual existing arguments. If you are doing exploratory research or building something new that does not yet have a debate around it, forcing a they say I say structure can feel artificial. In those cases, you are better off using a problem-solution frame or a classification structure instead. The framework is not a universal fix for every writing situation. For practical implementation, you do not need to memorize the entire book. The most useful section is the one on openings, because that is where almost everyone struggles. Start there. Write your source engagement sentences using the templates as scaffolding, then replace the templates with language that sounds like you. Check that your summary of the other position is accurate enough to survive scrutiny. Make sure your pivot actually introduces a new point instead of just repeating your thesis in different words. That last one is a common failure mode that I see constantly. The resource itself is available through their publisher's website and major academic retailers. The original title is They Say / I Say: The Moves That Moves in Academic Writing, and there are adapted versions for different disciplines and high school level users. I recommend the original for college and graduate work unless your instructor specifies otherwise.