How to Synthesize Sources in a Literature Review: Beyond Summarizing

Synthesizing is not summarizing. A summary reports what one source says. A synthesis combines multiple sources around a theme and adds your own analytical voice. If you can rearrange your paragraph in a different order—Study A, then Study B, then Study C—and it still makes perfect sense, you are summarizing. If your paragraph would fall apart without every study linked under one idea, you have synthesized.

This guide teaches you the paragraph-level mechanics of synthesis: the four synthesis moves that Tulane University identifies, signal phrases you can use immediately, the “They Say / I Say” formula for building analytical paragraphs, and a self-editing test that catches summary writing before your professor does.

You will learn how to move from a “shopping list” of sources to a paragraph that argues something about the field. The examples use the same remote work topic across multiple moves so you can see how the same set of sources transforms differently depending on which synthesis move you apply.

• Synthesis combines multiple sources around a theme and adds your analytical voice. Summary restates one source at a time.
• The four synthesis moves are Pattern (agreement), Debate (disagreement), Development (evolution), and Gap (what’s missing).
• Use the “rearrangement test” to self-edit: if your paragraph could be rearranged and still make sense, you’re summarizing.
• Start topic sentences with conceptual claims, not author names.
• Build a synthesis matrix before writing—let themes emerge from the data, not be guessed.
• Always include a gap statement to position your research within the field.

Summary vs. Synthesis at the Paragraph Level

The single most common feedback you will receive on a literature review is “this is too much summary” or “you need to synthesize more.” These comments sound vague because they often do. Here is exactly what they mean.

Summary reports what each author says, one at a time.

Martinez (2022) studied remote work productivity and found a 15% increase. Johnson (2021) examined work-from-home policies during COVID-19. Chen (2023) surveyed employee preferences and found 67% prefer hybrid models.

Synthesis combines multiple sources around a theme and adds your own analytical point.

Research reveals tension between productivity metrics and employee preferences in remote work arrangements. While Martinez (2022) documented a 15% productivity increase, Chen’s (2023) finding that 67% of workers prefer hybrid models suggests that pandemic-era policies (Johnson, 2021) may need refinement to balance organizational efficiency with worker autonomy.

The first paragraph tells you what each study found. The second paragraph tells you what those findings mean together. Your analytical voice guides the reader. This is what Dr. Theresa Orr, an academic writing expert, calls the difference between being a reporter and being a thinker: “A summary is when you take what you’ve read and reduce it down into your own words, focusing only on the main point. But when you synthesize you actually combine all the information together that you’ve reviewed to discover a new idea or a new point of view that is completely from you.”

Comparison Table: Summary vs. Synthesis

Feature	Summary	Synthesis
Definition	Shortens a single source into your own words, keeping only the main point.	Combines ideas from multiple sources to create a new point of view.
Purpose	Show you understand what one author is saying.	Show you can connect sources, spot patterns, and form your own argument.
What It Looks Like	One sentence per article, lined up one after another.	One paragraph that blends studies together under one clear idea.
Your Role	Reporter: you’re passing on what the author said.	Thinker: you add your own analysis of what the studies mean together.

The Johns Hopkins University synthesis matrix guide reinforces this distinction: “By arranging your sources by theme or variable, you can see how your sources relate to each other, and can start thinking about how you weave them together to create a narrative.”

The Four Synthesis Moves

Synthesis isn’t one thing—it’s a set of moves you make depending on what the sources are doing relative to each other. Tulane University’s library guide identifies the four most common moves. Each move requires different language and structure.

Move 1: Pattern (Agreement)

Use when multiple sources agree or converge on the same finding. You’re showing the reader that a consensus exists.

Signal phrases: “studies converge on,” “multiple sources agree,” “research consistently demonstrates,” “taken together”

Example:

Three recent studies on food insecurity — Okafor (2019), Lin & Patel (2021), and Reyes (2023) — all find that transportation access is a stronger predictor of food insecurity than household income alone. Why this move? Three independent studies reach the same finding. That convergence across sources is the pattern move — you’re showing the reader that a consensus exists on this point.

Patterns don’t require many sources. What matters is that multiple studies confirm the same finding. Bakke (2015) and Fernandez (2017) both document significant mental health benefits of green space access among urban youth. Two sources, but both pointing in the same direction, is still a pattern.

Move 2: Debate (Disagreement)

Use when sources disagree or take different positions. Don’t pretend they agree—the tension itself is valuable and shows the complexity of the field.

Signal phrases: “scholars disagree about,” “one camp argues,” “while X contends, Y maintains,” “conflicting evidence suggests”

Example:

Nguyen (2018) argues that community policing reduces violent crime, while Abrams & Cole (2020) contend that the same programs have no measurable effect and may actually erode community trust in law enforcement. Nguyen and Abrams & Cole reach opposite conclusions. The tension between them is the point — you’re not resolving it, you’re showing the reader that a genuine scholarly disagreement exists.

The Oxford guide’s dialectic synthesis framework adds depth here: “In contemporary academics, the thesis, antithesis, synthesis pattern has become a conventional approach to structuring academic syntheses, particularly where one source provides a counter-argument to another source. It is seen in all disciplines.” Following this method, knowledge is gained through proposing a thesis (a scholarly position), considering the antithesis (the reaction or negation), and arriving at a synthesis (a new idea stemming from the common elements the writer observes).

Move 3: Development (Evolution Over Time)

Use when you can show how thinking on a topic has changed over time. This move demonstrates your grasp of how a scholarly conversation has evolved.

Signal phrases: “early research assumed,” “more recent work has complicated,” “the field has shifted from asking whether to asking under what conditions,” “initially, scholars believed,” “subsequent research refined”

Example:

Early scholarship treated social media and political participation as straightforwardly linked (Harmon, 2010; Voss, 2012). By the late 2010s, researchers began questioning this, finding that online engagement often substitutes for rather than supplements offline action (Park, 2018; Delacroix, 2020). The passage shows a chronological arc — early assumptions giving way to more complicated findings. The key signal is the shift over time, not just that sources disagree.

Move 4: Gap (What’s Missing)

Use when you want to show what the existing literature hasn’t addressed. This is especially useful for positioning your own research or argument.

Signal phrases: “while X has been studied, few have addressed,” “despite extensive research on X, little attention has been paid to,” “what remains unexplored is,” “an important question still unanswered”

Example:

Although scholars have extensively examined the economic effects of gentrification on housing costs and displacement, very few studies have investigated how gentrification affects the social networks and mutual aid relationships of long-term residents. The passage acknowledges what has been studied (economic effects) then names what hasn’t (social networks). That “what’s missing” structure is the classic gap move.

Research on immigrant integration has focused heavily on economic outcomes such as employment and wages. Comparatively little attention has been paid to how integration policies shape civic participation and sense of political belonging. Like the previous gap example, this names what’s been studied versus what’s been overlooked. Two gap cards in the set is intentional—gaps are one of the most common and important synthesis moves in social science writing.

Signal Phrases Cheat Sheet

Here are ready-to-use signal phrases organized by synthesis move. You can copy these directly into your paragraphs.

Pattern Signals

• “Multiple studies confirm that…”
• “Research consistently demonstrates…”
• “Taken together, these findings suggest…”
• “Studies converge on the finding that…”
• “A clear pattern across the literature is…”

Debate Signals

• “Scholars disagree about…”
• “One camp argues that X, while others contend that Y.”
• “Conflicting evidence suggests…”
• “The debate over X centers on…”
• “While X maintains that Y, alternative explanations emphasize…”

Development Signals

• “Early research assumed…”
• “More recent work has complicated this view by showing…”
• “The field has shifted from asking whether to asking under what conditions.”
• “Initial findings suggested…, but subsequent research refined this to…”
• “What began as a focus on X has evolved toward an emphasis on…”

Gap Signals

• “While existing research has examined X and Y, few studies have addressed Z.”
• “Despite extensive attention on X, the impact of Y remains underexplored.”
• “An important question still unanswered is…”
• “Although X has been studied, little attention has been paid to…”
• “The literature lacks research on…”

The “They Say / I Say” Formula for Synthesis Paragraphs

The “They Say / I Say” formula was popularized by academic writing experts Graeter and Hartigan. Applied to synthesis paragraphs, it works like this:

1. Topic Sentence — States the analytical point (your “I Say”)
2. They Say — Combines multiple sources that support or illustrate that point
3. I Say — Your analytical voice explains what the sources collectively reveal

Weak example (summary style, no “I Say” voice):

Several studies have examined social media and teen mental health. Twenge et al. (2018) found higher social media use linked to depression. Fardouly et al. (2015) studied body image. Valkenburg and Peter (2013) examined online communication.

This is one study per sentence, no connection, no analytical voice.

Strong example (synthesis style with “I Say”):

Studies on social media and teen mental health show both concern and ambiguity. Twenge et al. (2018) found that heavy screen use correlates with rising depression in adolescents, while Valkenburg and Peter (2013) argue that supportive online communication improves social connectedness. Fardouly et al. (2015) focus on social comparison, showing that idealized images may harm body satisfaction. Together, these studies suggest that social media’s effects aren’t simply good or bad—they depend on how teens interact with online spaces.

The strong version identifies a pattern (platform-specific effects), explains what the pattern might mean (not a universal risk), and connects to a practical implication. It synthesizes four studies into one analytical claim.

The Rearrangement Test for Self-Editing

Here is Dr. Theresa Orr’s favorite way to check if you’ve actually synthesized instead of summarized:

If your paragraph could be rearranged without losing meaning (Study A, then Study B, then Study C), it’s probably just a summary. If your paragraph wouldn’t make sense without all the studies being linked under one idea, that’s a synthesis.

Apply this test before submitting:

1. Read each paragraph in reverse order — Paragraph 5, then Paragraph 4, then Paragraph 3. Does it still make logical sense?
2. Check for “shopping list” paragraphs — If every paragraph starts with a researcher’s name, you are describing rather than synthesizing.
3. Verify the topic sentence — Does it make a conceptual claim (not an author name)? If the topic sentence says “The shift to remote work has exposed tensions between…” that’s conceptual. If it says “Martinez (2022) studied…” that’s summary.
4. Check for signal phrases — Do your paragraphs use linking language (“although,” “however,” “converge,” “complicates”) to show relationships between sources?

Annotated Weak-to-Strong Examples

Let’s walk through how the same set of sources transforms under each synthesis move. The following sources represent illustrative studies on remote work and productivity—names and years are placeholders used to demonstrate form.

The Source Set

• Author A (2019): Remote work increases productivity (fewer interruptions, autonomy)
• Author B (2021): Productivity increases, then plateaus (home setup, role type matter)
• Author C (2021): Remote work decreases productivity (weaker team communication)
• Author D (2023): Mixed results (depends on task complexity)

Weak Version: Descriptive (Summary Style)

Author A (2019) found that remote work increased productivity. Author B (2021) also found an increase but said it levelled off. Author C (2021) found a decrease. Author D (2023) found mixed results.

This is a list. Each sentence stands alone, the sources do not interact, and no gap emerges.

Strong Version: Synthesised (Analytical Style)

Early evidence suggested that remote work raises productivity, largely by reducing interruptions and increasing autonomy (Author A, 2019). [Opens with a theme and groups the “increase” camp.] More recent work complicates this picture: while productivity gains appear in the short term, they may plateau as the novelty fades and depend heavily on the employee’s home setup and role (Author B, 2021). [Uses “complicates” to signal tension and introduces mediating factors.] A competing strand reports declines, attributing them to weaker team communication and coordination costs (Author C, 2021). [Names the opposing camp and its mechanism.] These apparently contradictory results may be reconciled by task complexity: remote work appears to benefit independent, focused tasks but to hinder highly interdependent ones (Author D, 2023). [Synthesises both camps under one explanatory factor.] Taken together, the literature has moved from asking whether remote work helps to asking under what conditions it does — yet few studies isolate task type and communication structure simultaneously, leaving this interaction underexplored. [States the trend and names the precise gap the new study can fill.]

Notice what happened: the second version is organised by a theme (not one paragraph per author), it makes the sources argue with each other, it uses linking language to signal relationships, and it ends by naming a specific, researchable gap. This is exactly the function a literature review is meant to serve.

Your Synthesis Matrix: The Tool That Makes It Possible

Before writing, you need a synthesis matrix—a table that arranges your sources by theme instead of one-by-one. The matrix reveals where studies agree, conflict, or leave gaps. Those rows-into-columns become your paragraphs.

Template Structure

Source (Author, Year)	Theme A (Productivity Effect)	Theme B (Mediating Factors)	Theme C (Methodology)	Relevance to My Question
Author A (2019)	Increases	Fewer interruptions, autonomy	Survey	Initial evidence for effect
Author B (2021)	Increases, then plateaus	Home setup, role type	Longitudinal	Complicates initial findings
Author C (2021)	Decreases	Weaker team communication	Case study	Conflicting evidence
Author D (2023)	Mixed	Task complexity, communication structure	Mixed-methods	Potential reconciliation

The matrix already reveals the structure: studies agree productivity changes, disagree on direction, and the disagreement seems to hinge on mediating factors. This insight would not be obvious if you were writing paragraph-by-paragraph summaries.

How to Build One

1. List your themes as columns — These emerge from your research question and initial reading. Don’t put one column per source—that creates the “shopping list” problem you’re trying to avoid.
2. List each source as a row — Include a brief citation: author name, year, and one-line study description.
3. Fill each cell with 3–5 bullet points — Capture what the source found or argued relevant to that theme. Keep it concise.
4. Look across the rows — When you see that the “Productivity Effect” column clusters around increases vs. decreases, that’s a theme you can structure your review around.
5. Look across the columns — When you see that the “Mediating Factors” column has entries for every source, that suggests mediation is a key pattern in your literature.
6. Spot the blanks — An empty cell in the matrix (a theme with no source entries) is itself a finding. It signals a research gap.

The University of North Dakota’s writing resources explains: “A synthesis matrix is an excellent tool to use to organize sources by theme and to be able to see the similarities and differences as well as any important patterns in the methodology and recommendations for future research.”

What to Avoid

• One paragraph per author — A sign of description; group studies that speak to the same point.
• No clear organising concept — Anchor your review to your research question.
• Ignoring disagreement — Include studies and findings that complicate the picture.
• Structure before research — Building a thematic structure without building a matrix first. Your themes should emerge from the matrix analysis, not be guessed at before reading the literature.

When to Choose Pattern vs. Debate vs. Development vs. Gap

The synthesis move you choose depends on what the sources are doing relative to each other. Here is a practical decision framework.

If the sources…	Use this move	Write a paragraph that…
Agree or converge on the same finding	Pattern	Opens with a consensus claim, cites multiple confirming studies, explains what the pattern means
Conflict or take different positions	Debate	Names the competing camps, cites representative studies for each, identifies what the disagreement centers on
Show thinking shifting over time	Development	Opens with the early view, traces how it changed, explains why the shift occurred
Leave a question unanswered	Gap	Acknowledges what’s been studied, names what’s missing, explains why the gap matters

Recommendation: Most literature reviews use a mix of all four moves within a single review. A typical structure might include: two Pattern sections (establishing consensus), one Debate section (mapping disagreement), one Development section (tracing evolution), and a Gap section (positioning your research). The exact mix depends on your topic.

Tradeoff: If your topic is newly emerging and few studies exist, you may lean heavily on Gap and Pattern. If your topic has a long research history, you may need more Development and Debate. The key is matching the move to the source relationships—not forcing a move that doesn’t fit.

What We Recommend

Here is what most students should do, and why:

Start with the synthesis matrix, not the outline. Your outline should emerge from the matrix, not be guessed before reading the literature. This reverses the default student approach (outline first, matrix never) and makes the actual synthesis work visible before you write a single paragraph.

Write topic sentences that make conceptual claims, not author lists. Every paragraph should start with an idea (“The shift to remote work has exposed tensions between…”) not an author (“Martinez (2022) studied…”). Your analytical voice should guide the reader.

Use the “rearrangement test” before submitting. If your paragraph could be rearranged without losing meaning, it’s probably just a summary. If your paragraph would fall apart without every study linked under one idea, you have synthesized.

Include a gap statement in every literature review. Even if your assignment doesn’t require it, naming what the literature hasn’t addressed shows you understand the field—and examiners notice.

Common Mistakes to Warn About

1. Summarising instead of synthesising — Organise by theme, not by source. If every paragraph begins with a researcher’s name, you are describing rather than synthesising.

2. One paragraph per author — A sign of description; group studies that speak to the same point. A synthesis paragraph should weave together two or more sources.

3. No clear organising concept — Anchor the review to your research question. Without a conceptual anchor, paragraphs drift into summary.

4. Ignoring disagreement — Include studies and findings that complicate the picture. A review that only reports consensus is incomplete.

5. Citation errors and accidental plagiarism — Track sources carefully and cite consistently. This is not just about academic integrity—it’s about showing your reader exactly which source supports which claim.

6. Trusting AI-generated references — General-purpose chatbots can fabricate citations and misread findings. Any AI output must be verified against the original paper. Use AI to accelerate the mechanical work; keep the analysis your own.

Your Next Steps

1. Build your synthesis matrix — Organize your most relevant sources by theme, methodology, and findings.
2. Apply the four synthesis moves — For each theme, decide which move (Pattern, Debate, Development, Gap) best describes the source relationships.
3. Use the signal phrases — Copy the ready-to-use phrases above into your draft paragraphs.
4. Run the rearrangement test — Read your paragraphs in reverse order. If they still make sense, you’re summarizing.
5. Write topic sentences as conceptual claims — Not author names. Ideas.

Further Reading

If you want to go deeper:

• Synthesis matrix templates for all academic levels — Our guide on How to Write a Literature Review: Templates and Synthesis Matrix shows how to build and use a synthesis matrix at undergraduate, graduate, and PhD levels.
• Advanced synthesis techniques for thesis writing — Our guide on How to Write a Critical Literature Review for a Thesis: Advanced Synthesis Techniques covers the 5 C’s framework (Cite, Compare, Contrast, Critique, Connect) and gap identification strategies.
• Literature review introduction structure — Our guide on How to Write a Literature Review Introduction: 5-Part Structure with Examples walks through the exact opening structure for orienting readers and justifying your thesis.