Secondary referencing is what you do when the source you want to cite is one you have only seen quoted, summarised or discussed inside another source — you are citing it second-hand, through the book or article you actually read. Harvard handles this honestly and neatly: the words “cited in” appear in your in-text citation, naming both the original author and the author of the source in front of you, and your reference list contains only the source you genuinely read. Get either half of that wrong and you are, in effect, claiming reading you have not done, which is why markers treat this topic more seriously than its small word count suggests. This guide explains when secondary referencing is legitimate, the exact in-text format, the reference list rule, and how to escape the situation by finding the original. If you are building a full Harvard list, the Harvard Referencing Generator formats your primary references, and our guides to Harvard in-text citations and the collected Harvard referencing examples cover the everyday source types this one sits alongside.
What secondary referencing actually is
A secondary reference has two layers. The original (primary) source is the work where the idea, finding or quotation first appeared. The secondary source is the work you are holding — the textbook, journal article or chapter that quotes or discusses the original. When you write that a researcher found something, but your knowledge of that finding comes entirely from a later author’s account of it, you are relying on a chain: original author, then the intermediary who selected and framed it, then you. Secondary referencing exists to make that chain visible to your reader rather than hiding it.
The distinction matters because an intermediary is never a neutral pipe. The author you read chose which sentence of the original to quote, which result to report and which qualifications to leave out. Their summary may be accurate, but it is still their interpretation, written for their argument. When you cite the original directly as though you had read it, you inherit any distortion invisibly. When you cite it with “cited in”, you tell the marker exactly how far your reading goes — and that honesty is precisely what the format is designed to reward.
When it is acceptable — and when markers read it as lazy
Secondary referencing is acceptable when the original is genuinely out of reach despite a real attempt to find it. The recognised cases are consistent across UK university guides: the original is out of print and not held by your library; it is not published in English and no translation exists that you can use; it is an obscure or very old work, a conference paper never formally published, or a source held only in an archive you cannot visit. In each case the common thread is that a reasonable search failed, not that a search was never made.
Markers read secondary referencing as lazy in the mirror-image situations. If the “original” is a famous, widely held text — a standard textbook, a classic study reprinted in every collection, a book sitting in your own library — then “cited in” announces that you did not do the basic work of fetching it. It looks worst of all when the borrowed source is load-bearing: if your argument depends on a finding, your marker expects you to have read that finding in its original context, with its methods and caveats attached. A scattering of secondary citations for passing mentions is unremarkable. An essay whose key evidence is all second-hand reads as an essay written from one textbook’s reading list, and it is marked accordingly.
The in-text format: original, year, cited in, secondary, year, page
The Harvard in-text format names both works in a fixed order: the original author’s surname and the original year, then the words cited in, then the secondary author’s surname, the secondary year, and the page of the secondary source where you found the material. In parenthetical form:
(Halloway, 1987, cited in Mercer, 2019, p. 203)
Here Halloway’s 1987 study is the original you have not read; Mercer’s 2019 book is the one open on your desk, and page 203 is where Mercer reports Halloway. That page number always belongs to the secondary source, because that is the page your reader can actually check. In narrative form, where the original author is named in your sentence, the structure follows the same order:
Halloway (1987, cited in Mercer, 2019, p. 203) reported that trainee nurses described supervision as “the difference between coping and quitting”.
Two details separate correct Harvard from near-misses. First, Harvard uses “cited in” — not “as cited in”, which is the APA phrasing; students who move between styles import the wrong phrase constantly. Second, both years appear. Some styles drop the original year, but Cite Them Right Harvard keeps it, because the date of the original is part of what you are telling the reader: a 1987 finding reported in 2019 is a different claim from a recent one. If you are quoting the original author’s exact words as reproduced in the secondary source, the same citation applies — the quotation marks belong to the original words, but the page number and the “cited in” route still point to where you found them.
The reference list rule: only the source you actually read
This is the rule students break most often, and it is absolute in Cite Them Right Harvard: your reference list contains the secondary source only. The original work does not get an entry, because a reference list is a record of what you read, and you did not read it. The in-text citation carries the original author’s name so the reader knows whose idea it was; the list carries the book you can be examined on.
So for the example above, the list entry is Mercer’s book, in full normal Harvard book format — and there is no entry for Halloway at all:
Mercer, J. (2019) Learning on the ward: a guide for nursing mentors. London: Sage.
Never construct a reference list entry for the original from the fragments the secondary source gives you. You do not know its full title with certainty, its publisher or its edition, and a fabricated-looking entry for a book you never opened is a worse offence than the secondary citation itself — it converts an honest shortcut into a misrepresentation. If your marker wants to trace the original, the “cited in” citation plus the secondary source’s own reference list gives them the trail.
Worked examples: book in a book, article in a textbook
The commonest real case is an older book discussed inside a newer one. Suppose your textbook on study skills quotes a 1970s education researcher on note-taking. You cite the claim like this:
Students who reorganise their notes within a day retain substantially more than students who simply re-read them (Braddock, 1974, cited in Fenner, 2021, p. 88).
Fenner, L. (2021) Studying at university: skills for success. 3rd edn. Basingstoke: Palgrave Macmillan.
The second common case is a journal article you know only through a textbook’s account of it — typical where the article sits behind a paywall your library does not cover, in a journal your institution does not subscribe to:
Ward-based assessments were rated fairer by students than end-of-placement examinations (Okafor, 2011, cited in Mercer, 2019, p. 156).
The reference list again carries only Mercer (2019), exactly as in the entry above. Notice what does not change between the two examples: the pattern is identical whether the original was a book, an article or a report, because the format describes the relationship between the sources, not their types. For the standard format of the textbook itself in your list, our guide on how to reference a book in Harvard style gives the full rules.
Chaining: why one link is the limit
Occasionally you will meet a source that is itself citing second-hand — your textbook reports Author B, who was reporting Author A. It is tempting to reproduce the whole chain, but Harvard secondary referencing supports one link only: you cite what your immediate source claims, attributed through that source. If the chain matters, cite your source’s own statement about it, or better, treat the tangle as a signal to find the original.
There is a sound reason for the limit beyond tidiness. Each retelling in a chain is a fresh act of selection and compression — the academic version of a whispered message passing down a line. By the third retelling, figures get rounded, conditions fall away and tentative findings harden into facts. A citation that honestly says “A, cited in B” already carries one layer of that risk, which is why it is reserved for minor points. A claim that has passed through two intermediaries before reaching you is not evidence you should build on at all; it is a lead to follow back to the source.
Finding the original instead: the search to make first
Before you settle for “cited in”, run the search a marker would expect. Start with your university library catalogue, including its e-book collections — many older works exist as scanned e-books even when the print copy has vanished from the shelves. Search the title in a general web search and in Google Books or a similar preview service; even a partial preview may give you the exact passage you need, at which point you have read the original for that passage and can cite it directly, honestly. Check whether the work was reprinted or revised: classic studies are frequently republished in readers, collected editions and anniversary editions under the same title with a new date, and citing the reprint you actually read is a primary citation, not a secondary one.
If the item exists but your library does not hold it, use the inter-library loan service — most UK university libraries will obtain a book or a scanned journal article from another institution, usually within days, and a chapter or article scan is often faster than a full loan. Only when these routes fail — the loan cannot be fulfilled, the work is untranslated, the item is genuinely untraceable — is the secondary citation the correct professional answer rather than a shortcut. Keeping a brief note of what you tried is worth doing: if a tutor questions the citation, “requested through inter-library loan on this date, unavailable” closes the conversation.
The errors that give secondary referencing a bad name
Four errors account for nearly all the marks lost here. Citing the original as if read — writing (Halloway, 1987) with a full list entry built from guesswork — is the serious one, a small act of academic dishonesty. Listing both sources in the reference list is the confused one: it tells the marker you do not know the rule. Omitting the page number of the secondary source makes the citation unverifiable, since the whole point is that the reader can find the passage in the book you read. And overuse — a “cited in” in every paragraph — converts a legitimate device into a pattern that undermines the essay’s claim to be based on reading at all. Used rarely, precisely, and only where the original is truly out of reach, secondary referencing costs you nothing. Used as a method, it costs you credibility.
Struggling to keep your Harvard references honest and accurate?
If your source list has grown chains of second-hand citations, or you are unsure which entries belong in your reference list at all, United Academic Help can help you straighten the referencing out alongside the writing itself. Our editors and writers support essays, coursework and dissertations with referencing and writing help from £20 per 1,000 words, and the full price is agreed with you before any work begins — no surprises afterwards. Place an order through the website, email your brief to team@unitedacademichelp.uk, or reach us on WhatsApp via our contact page and tell us which style guide your department uses.
Frequently asked questions
Do I put both sources in my Harvard reference list? No. Only the source you actually read — the secondary source — appears in the reference list. The original author is named in the in-text “cited in” citation only.
Which page number do I give in a secondary citation? The page of the secondary source, where you found the material reported. That is the page your reader can check, since the original is the work neither of you may be able to open.
Is it “cited in” or “as cited in” in Harvard? Harvard uses “cited in”. “As cited in” is the APA form. Mixing the two is a common slip when a course uses different styles in different modules.
How many secondary references can I use in one essay? There is no fixed number, but they should be rare and never carry your central argument. If a source matters to your thesis, find and read the original; keep “cited in” for passing mentions of genuinely unavailable works.
What if the original is in another language? If no translation exists and you cannot read the original, a secondary citation through an English-language source that discusses it is acceptable. Say what you relied on; do not cite the foreign-language original as though you had read it.
Can I quote the original author’s exact words through a secondary source? Yes, if the secondary source reproduces them as a quotation. Keep the quotation marks, cite with “cited in” and the secondary source’s page, and list only the secondary source.