Live Chat
Here is one journal article, annotated twice — once in APA 7 and once in MLA 9 — so you can see exactly what changes and what does not.
Hartman, S. (2008). Venus in two acts. Small Axe, 12(2), 1–14. https://doi.org/10.1215/-12-2-1
Hartman reads the archive of Atlantic slavery through one legal document — an indictment of a slave-ship captain for the murder of two girls — and argues that the record preserves the violence that produced it, so recovering a life from it is never neutral. Her method is what she names critical fabulation: writing from the gap rather than filling it. Its value for this paper is that vocabulary, which lets me treat silence as something the archive does rather than something it lacks. The essay works from one fragment and says so, which makes it a method and a caution, not a body of evidence.
Hartman, Saidiya. "Venus in Two Acts." Small Axe, vol. 12, no. 2, 2008, pp. 1–14. https://doi.org/10.1215/-12-2-1.
Hartman reads the archive of Atlantic slavery through one legal document — an indictment of a slave-ship captain for the murder of two girls — and argues that the record preserves the violence that produced it, so recovering a life from it is never neutral. Her method is what she names critical fabulation: writing from the gap rather than filling it. Its value for this paper is that vocabulary, which lets me treat silence as something the archive does rather than something it lacks. The essay works from one fragment and says so, which makes it a method and a caution, not a body of evidence.
The annotation is word for word the same in both. What changes is the citation above it and the indent underneath it — that is the whole of the difference, and it is why annotated bibliographies turn out easier than most students expect.
Both sources are real and checkable — Saidiya Hartman's essay in Small Axe and Michel-Rolph Trouillot's Silencing the Past — so you can look the entries up and confirm every element. The annotations are written as a student would write them, for one imagined project on how archives record silence. Read the sources yourself before citing either.
- An annotation is not a summary. It summarizes, then evaluates, then says why the source matters to your project.
- APA: annotation is a new paragraph under the reference, the whole block indented 0.5 in.
- MLA: annotation starts on a new line after the entry, indented one inch, so it is visibly deeper than the hanging indent.
- Chicago: notes-bibliography entry, annotation indented 0.5 in — and spacing is the one thing you must check locally.
- All three order entries alphabetically by author surname, exactly as a normal reference list.
- Typical length is 100–200 words per annotation unless your brief says otherwise.
- Writing in the first person about your own project is normal here, and unusual almost everywhere else.
The three jobs every annotation does
A weak annotation does only the first of these. A strong one does all three, usually in this order.
What is the source's argument, and what evidence is behind it? Two or three sentences. Not a blurb — the actual claim.
Is it credible? What is the method, the sample, the date, the obvious limitation? This is where most annotations stop short, and where most of the credit sits.
What is it doing in your bibliography? Which part of your argument does it support, complicate, or contradict?
Step 3 turns a list of sources into a piece of thinking. It is also why an annotated bibliography cannot be handed to a citation generator: the generator does not know what your paper is about.
How to spend the words
Most annotations run out of room inside the summary and reach step 3 with nothing left. A rough split fixes that. Working targets, not rules from any manual:
| Assigned length | Summarize | Evaluate | Position |
|---|---|---|---|
| ~75 words | 2 sentences | 1 sentence | 1 sentence |
| ~150 words | 2–3 sentences | 2 sentences | 1–2 sentences |
| ~250 words | 3 sentences | 3–4 sentences | 2–3 sentences |
The sentence map
Read the APA annotation at the top of this page one sentence at a time. Each has a job you can copy.
| Sentence | Its job | In the example above |
|---|---|---|
| 1 | The claim, plus the evidence base | "argues that the record preserves the violence that produced it" |
| 2 | The method | "what she names critical fabulation" |
| 3 | What it does in your paper | "lets me treat silence as something the archive does" |
| 4 | The limit | "works from one fragment and says so" |
The order can move; the four jobs cannot be dropped. They map onto the three moves above — the claim is the summary, method and limit are the evaluation, use is the positioning — and every sentence you write should be doing at least one of them.
Present tense for what the source argues — "Hartman reads", "the book traces". Past tense for what its researchers did — "surveyed 400 undergraduates". Mixing the two is the most common tense error here, and it turns invisible once written.
Which of the three your brief is asking for
Purdue OWL sets out the same three moves — summarize, assess and reflect — and says an assignment may want one, two or all three. The verb in your prompt decides it.
| The verb in your brief | What the annotation should do |
|---|---|
| "Summarize your sources" | Descriptive. Words go to the argument and its evidence, with one closing line on relevance. |
| "Evaluate the usefulness or reliability" | Critical. Method, sample, date and the author's standing carry it. Name a limitation in every entry, including the sources you like. |
| "Explain how each source contributes to your project" | Reflective. Positioning is the bulk: a named section, a claim, a counter-argument. |
| "Annotate", or "summarize and evaluate" | All three — the default in most US courses, and what the samples here show. |
If the prompt names a framework such as CRAAP, mirror its vocabulary: a rubric that asks about "authority" wants that word in your evaluation sentence.
A weak annotation, a better one, and a strong one
Same source, three attempts, very different grades.
Weak
- "This article talks about Venus and the archive of slavery."
- "The author discusses several documents and makes interesting points about absence."
- "It was useful for my research."
Strong
- Names the actual argument, not the topic
- Names the method and its limit — one document, read closely
- Says which section of your paper it feeds and why
The weak version could have been written from the abstract in ninety seconds. That is precisely what a grader is checking for.
Between those poles sits the version most students actually submit.
Hartman reads the archive of Atlantic slavery through one legal document and argues that the record preserves the violence that made it. Her method is close reading of a single fragment rather than a survey, so the findings are interpretive. Working from one document limits how far the argument travels on its own evidence.
Accurate and genuinely evaluative — and this is where most annotations stop. Nothing in it says why this source is on your list rather than another.
Hartman reads the archive of Atlantic slavery through one legal document and argues that the record preserves the violence that made it, so recovery is never neutral. Her method — critical fabulation, her term — is close reading of a single fragment, which makes the essay interpretive rather than evidential. That is what this paper needs: a vocabulary for treating silence as an act rather than a gap, and it will carry my second chapter. It also sits against Trouillot, who locates silence in a describable process; their disagreement is about method, not evidence.
Argument, method, limit, use — then the move almost nobody makes: a sentence tying this entry to another on the same list. Annotations that talk to each other read as research.
A second entry: a book, in both styles
Journal articles are the easy case. A book changes the shape of the citation: the publisher arrives, the volume and page range disappear, and the year moves.
Trouillot, M.-R. (1995). Silencing the past: Power and the production of history. Beacon Press.
Trouillot argues that silences in the historical record are products of power rather than accidents, and that they enter at four moments: when a fact is created, when facts are assembled into an archive, when they are retrieved as a narrative, and when a period is given retrospective significance. He builds the case from extended examples — the Haitian Revolution, the Alamo, Columbus — rather than a survey of archives. I use the four-moment framework in chapter two, to say where in the process a particular silence was produced. Because the model comes out of a few cases, it is a lens to think with rather than a tested finding.
Trouillot, Michel-Rolph. Silencing the Past: Power and the Production of History. Beacon Press, 1995.
The annotation is left off the MLA entry because it does not change — same paragraph, moved out to a one-inch indent. In the citation itself, APA puts the year in parentheses straight after the author, cuts the given name to initials — a hyphenated first name keeps its hyphen, so Michel-Rolph becomes M.-R. — and uses sentence case for the title. MLA spells the name out, puts the year at the end of a book entry, and uses title case. Our breakdown of APA 7th against MLA 9th covers the rest of the entry types.
The third style: the same sources in Chicago
Chicago is what US history and art history courses ask for, and most example pages skip it. It runs two systems: notes-bibliography, ending in a page headed Bibliography, and author-date, ending in a reference list. Annotated bibliographies normally use notes-bibliography.
Trouillot, Michel-Rolph. Silencing the Past: Power and the Production of History. Beacon Press, 1995.
Trouillot locates archival silence at four moments in the making of history, from the creation of a single fact to its retrospective significance. His claim is that those silences are produced by power, not left behind by accident. The argument is built from three extended cases rather than a survey, so it reads as a framework rather than a result. Used here for the four-moment vocabulary, and set against Hartman, who doubts that naming the mechanism repairs anything.
That one is written to the 75-word row of the table above: two sentences of summary, one of evaluation, one of positioning. Same book, same judgment, compressed.
Hartman, Saidiya. "Venus in Two Acts." Small Axe 12, no. 2 (2008): 1–14. https://doi.org/10.1215/-12-2-1.
Its annotation would be the APA one from the top of this page, unchanged. Switching style is a citation problem, not a writing problem.
- In the article entry, volume and issue read "12, no. 2", the year sits in parentheses, and a colon introduces the page range. MLA uses commas throughout.
- The book entry no longer needs a place of publication. The Chicago Manual's notes-bibliography quick guide for the 18th edition says "a place of publication is no longer required in book citations". A 17th-edition entry read "Boston: Beacon Press, 1995", so an older handout will disagree with a current one.
- One consequence is worth noticing. With the place gone, a plain single-author book — no edition, no translator, no series — comes out identical in MLA 9 and Chicago 18. Compare the two Trouillot entries above. These styles part company on articles, not books.
APA and MLA are double-spaced throughout, and the guides agree. Chicago is where advice splits: some university guides double-space with no blank line between entries, while Turabian practice single-spaces each entry and leaves a blank line between them. Follow your department's model document and stay consistent down the page.
Formatting the page itself
The page is closer to a reference list than to an essay. Center the title — "Annotated Bibliography", bold in APA and plain in MLA, and simply "Bibliography" in Chicago. In APA and MLA, run the entries down the page with no blank lines between them. Inside each entry the citation keeps its ordinary hanging indent, and the annotation sits beneath it as a block, indented as a whole, with no extra indent on its first line.
Across all three styles at once:
| What to set | APA 7 | MLA 9 | Chicago 18 |
|---|---|---|---|
| Page title | Annotated Bibliography, bold | Annotated Bibliography | Bibliography |
| Order | Author surname | Author surname | Author surname |
| Annotation indent | 0.5 in | 1 in | 0.5 in |
| Spacing | Double | Double | Guides differ — check |
| Where the year sits | Parentheses, after the author | End of the entry in a book; before the page range in an article | After the publisher; in parentheses for articles |
| Title capitalization | Sentence case | Title case | Title case |
The indent figures come from Purdue OWL's annotated bibliography samples, which states that "MLA 9 now requires a 1-inch indent; APA and Chicago require a 0.5-inch indent." APA covers annotated bibliographies in section 9.51 of the Publication Manual and publishes a free student template you can type straight into.
If you find an MLA page describing a line space and then an ordinary paragraph indent instead of a one-inch block, check its date. The MLA Style Center post that says so was written for the eighth edition in 2016 and carries MLA's own notice pointing readers to the ninth. Use the one-inch version, or your instructor's model document if they gave you one.
Annotation length is set by your instructor far more often than by the style manual. Some ask for 75 words, some for a full 250 with a required evaluation sentence. If a word count is given anywhere in the assignment, it overrides the 100–200 range above.
If this bibliography sits inside a longer APA assignment, our APA research paper examples show what the pages around it should look like.
Sources that do not look like a journal article
Most library guides show an article and a book. Real reading lists are messier, and the awkward entries are where points get lost.
A web page with no named author. Alphabetize by the organization, or by the title if there is none. The annotation carries the authority work: who runs or funds the site, and whether the page cites anything at all.
A primary source — a novel, a photograph, interview transcripts, a dataset. Summarizing becomes describing: what it is, how much of it, where it is held. Evaluating becomes provenance: who made it, when, and for whom.
A news article. Give the outlet, the date, and whether the piece is reporting or opinion. Annotating a news feature as though it were a study loses points faster than anything else here.
A source you decided against. Keep it in and say why it is out — superseded, weak for your question, or about a different population. A reasoned rejection reads as judgment.
An older source. Age is not a fault. If a 1994 article set the terms the field still argues in, say so. If its data has been superseded, name what replaced it.
A preprint or unpublished thesis. State that it has not been peer-reviewed, in plain words. That sentence is the evaluation step doing its job.
What goes wrong, and how to fix it
| The problem | How it shows up | The fix |
|---|---|---|
| The abstract in disguise | Phrasing that echoes the source's own abstract | Close the abstract; write from your own notes on the full text |
| Topic instead of claim | "This article is about silence in the archive" | Write what the author argues that someone could dispute |
| No positioning | The annotation ends on evaluation | Add a sentence starting "For this paper…" naming a section or claim |
| Generator output pasted in | Wrong capitalization, missing italics, "n.d." with a date available | Check by eye — generators are reliable on field order, not capitalization |
| Mixed styles down the page | APA punctuation in an MLA entry; two indent depths | Format one entry correctly, then copy its shape |
How many sources, and how to choose them
The count comes from the brief. What the brief will not tell you is that the sources should disagree with each other.
Ten sources that all say the same thing show that you found a consensus and stopped. Naming the source that cuts against your reading — and saying what the disagreement is actually about — is what separates a research bibliography from a reading list.
If you are still choosing sources, our guidance on keeping research original covers how to record what you take from each one as you read.
When the bibliography is the assignment
Sometimes it is standalone coursework; sometimes it is a graded checkpoint before a research paper, and that decides how much you invest in step 3. As a checkpoint, your positioning sentences are the outline of a paper you have not written yet, and your instructor is reading them to see whether it will work. If that paper comes next, our step-by-step guide to writing a research paper picks up from here.
A last pass before you submit
Read the page once for each line below. One pass per check beats one read looking for everything.
- Every entry is in one style, and it is the style the brief names.
- Every annotation names an argument, not a topic.
- Every annotation has one evaluative sentence a reader could disagree with, and one saying what the source does in your paper.
- Present tense for what the source argues; past tense for what its researchers did.
- No annotation reuses a phrase from the source's own abstract.
- Indents are uniform: 0.5 in for APA and Chicago, 1 in for MLA.
- Spacing is uniform down the page, inside annotations as well as between entries.
- Each entry sits inside whatever word range the brief gave.
If the source list is long and the deadline is not, our annotated bibliography service builds entries to your required style with the evaluation and positioning written for your specific project, from $7.89 a page. It sits inside our wider paper writing service for the research paper that usually follows.
