How to cite a source you found in another source

You are reading Smith (2019). Smith quotes a striking line from Jones (1987). The line is exactly what your argument needs.

You have not read Jones. You have read Smith reading Jones.

That distinction is what a secondary citation exists to record, and every major style has a form for it — because the alternative, citing Jones directly as though you had read it, is a small dishonesty that occasionally becomes a large one.

The formats

The pattern is identical across all four: name both in the text, list only the one you read. The reference list is a record of what you consulted, and you did not consult Jones.

Why only the source you read goes in the list

Because the list is a set of instructions for finding what you used. A reader who wants to check your claim needs to reach Smith, page 214, where the quotation appears in the form you saw it.

Putting Jones in the list claims you held it, which is a claim you cannot support. It also breaks the one-to-one relationship a reference list depends on — see our guide on reference lists and bibliographies for why that matters more than it sounds.

Why you should usually find the original instead

Secondary citations are legitimate, and markers accept them. They are also a signal, and what they signal is that you worked from summaries.

Two concrete reasons to go and find Jones:

Misquotation propagates. A claim gets slightly reshaped each time it is relayed, and by the third relay it can say something the original did not. This is not hypothetical — there are well-known cases in several fields of a finding being cited for decades in a form its authors never claimed. When you cite Jones through Smith, you are vouching for Smith’s reading of Jones, and you cannot check it.

The context is usually load-bearing. Smith quoted the line because it served Smith’s argument. Jones may have written it as a caveat, an aside, or a position later abandoned in the same paper. You cannot know which without looking.

When a secondary citation is genuinely right

Not every case needs chasing down. It is reasonable when:

• The original is unobtainable — out of print, in an archive, in a language you cannot read, or behind a paywall your institution does not have.
• The original is not the point. If your argument is about how Smith uses Jones — a historiographical or literature-review claim — then Smith is the source, and you are citing Jones only to identify what Smith was working with.
• It is a passing reference. A single background sentence does not warrant an inter-library loan.

The test: would your argument change if Smith had misread Jones? If yes, find Jones. If no, cite it secondarily and move on.

How to find the original quickly

Usually faster than students expect.

1. Smith’s reference list has the full details. That is what it is for. Copy them.
2. Search the exact title, in quotation marks, in your library catalogue or Google Scholar.
3. If it has a DOI, that resolves straight to it — and a DOI you can resolve is proof the source exists in the form claimed. Our citation generator will fetch the full details from a DOI, which turns “find the original” into a ten-second job once you have the number.
4. Inter-library loan for anything genuinely unavailable. It is free at most institutions and usually takes days rather than weeks.

When the intermediary is a textbook or lecture slides

The commonest secondary citation of all, and the one students least often recognise as one.

A textbook says “Vygotsky argued that learning is fundamentally social.” A lecture slide says “Bowlby (1969): attachment is an evolved survival mechanism.” Neither is Vygotsky or Bowlby. If you cite them directly you are claiming to have read work you met in a summary, and with foundational theorists that claim is unusually easy to test — a marker who knows the field can often tell from which phrasing you reproduce.

With canonical theorists there is a strong argument for finding the original anyway, because they are almost always easy to get and because textbook summaries of them are famously lossy. Vygotsky and Bowlby are both in most university libraries and substantial extracts are freely available; the inter-library loan case does not apply.

Lecture slides are a special case. Before citing anything through them, check whether your department permits citing lectures at all — many do not, which means the slide cannot be your intermediary and you have to reach the original regardless.

When Smith gives you nothing to go on

Occasionally the trail runs cold: Smith quotes Jones but gives an incomplete reference, or the reference list entry is wrong. This is more common than it should be.

Three things usually recover it:

1. Search the exact quoted phrase in quotation marks in Google Scholar or Google Books. A distinctive sentence of eight or more words almost always surfaces its source, and Google Books will often show you the page.
2. Check who else cites it. Look at other papers on the topic and see how they reference the same claim. If three of them agree on a year and a title that Smith got wrong, you have your answer and a small insight into Smith.
3. Ask a librarian. Genuinely — tracing a mangled reference is a routine part of the job, and subject librarians are both good at it and rarely asked.

If it truly cannot be found, say so. (Jones, n.d., as cited in Smith, 2019, p. 214; full details not traceable) is honest, and far better than inventing a plausible year.

A word on how many is too many

There is no rule, but there is a threshold at which a marker notices.

A bibliography where several key claims are all “as cited in” reads as an essay built from review articles rather than from sources. One or two, for genuinely hard-to-reach material, reads as diligence. The difference is visible, and it is the kind of thing that separates a 2:1 from a First.

Recording it as you go

The practical fix is a note-taking habit rather than a citation trick: when you copy a quotation out of Smith, record at the moment you copy it that it came from Jones via Smith, with both page numbers. Three weeks later you will not remember, and the failure mode is citing Jones directly by accident — which is the thing this whole page exists to avoid.

Our citation generator saves sources in your browser as you add them, so the source you captured in week two is still there in week nine with its details intact.