Secondary citations: when you cannot find the original

What a secondary citation actually claims

Referencing is a claim about what you read. When you write (Vygotsky, 1978) you are telling your marker you have Vygotsky in front of you. When you write (Vygotsky, 1978, cited in Wertsch, 1985) you are telling them something more honest and more limited: that Wertsch reports Vygotsky as saying this, and that you have read Wertsch.

That distinction is why the reference list contains only Wertsch. A reference list is a retrieval aid — it lists what a reader needs to check your work. To check your claim, a reader needs Wertsch's page, because that is where the report you relied on appears. Putting Vygotsky in the list would imply you consulted an edition you never opened.

Students often invert this, listing the original and omitting the intermediary. It looks tidier and it is the one version that is definitively wrong: it attributes to you a reading you did not do, and it removes the only source a reader could use to verify the quotation.

The format in each style

The wording differs slightly and the placement does not. In every case the original comes first, the intermediary second, and the intermediary alone goes in the list.

Name the original in your sentence

A secondary citation reads badly in parentheses and well in prose, because the sentence can carry the attribution that the bracket has to compress. "Wertsch reports Vygotsky's claim that the zone of proximal development is defined by what a learner can do with support (Wertsch, 1985, p. 45)" is clearer than the same point buried in a double citation.

This also protects you from the most common failure of secondary citing, which is quietly adopting the intermediary's interpretation as if it were the original's words. Wertsch is summarising, selecting and translating. Saying so is not a weakness in your writing; it is the accuracy the marker is checking for.

Where the intermediary is explicitly disagreeing with the original, say that too. "Wertsch reads Vygotsky as arguing X, though he notes the 1934 text is ambiguous on the point" is a sentence that demonstrates you understood what you were reading.

The four times it is genuinely acceptable

Markers penalise secondary citation when it stands in for reading you could have done. These four cases are not that.

• The original is out of print, unpublished, or held in one archive on another continent.
• The original is in a language you do not read and no translation exists.
• The original is a personal communication, an internal report, or data the intermediary holds.
• You are writing about the reception of an idea — how Wertsch read Vygotsky is your actual subject.

How to find the original before giving up

Most students reach for a secondary citation about four minutes before the original would have appeared. The intermediary's own reference list gives you the full record, and from there three routes usually work.

Search your library discovery layer by the exact title rather than the author, because author fields are inconsistently indexed for older works and translations. If nothing appears, try the title in Google Scholar and follow the 'All versions' link, which surfaces institutional repository copies that the publisher's page does not.

Where the work is genuinely unavailable to you, inter-library loan will usually get a chapter scan within a week, which is inside the timeframe of most coursework if you start early. For a dissertation there is no excuse at all: you have months.

The trap in the middle: transcription errors

Quotations degrade as they are passed along. A phrase gets trimmed, a hedge disappears, an emphasis is added. By the third repetition an author is being quoted as saying something firmer than they wrote.

This has real consequences in the disciplines where it happens most. Psychology and education have several famous claims that circulate in a form no original text supports, kept alive entirely by secondary citation from one textbook to the next.

So when you do cite secondarily, quote the intermediary's wording exactly and do not tidy it. If the phrasing looks odd, that is information — it may be a translation artefact, and smoothing it out would be inventing a text.

Secondary citation and word limits

A secondary citation costs more words than a direct one, and in a tight assignment that is a real pressure toward doing it badly. Resist the shortcut of dropping the intermediary to save four words; it converts an honest limitation into a false claim.

If the double citation appears three or four times, restructure instead. Attribute once in prose — "the account below follows Wertsch's reading of Vygotsky" — and then cite Wertsch normally throughout. That is accurate, shorter, and reads better.

What markers actually deduct for

Almost no rubric penalises a correctly formatted secondary citation on its own. What loses marks is the pattern around it.

• A reference list where the seminal works of the field appear only through intermediaries.
• Secondary citation of a source that is in your library, one floor up.
• The original in the reference list and the intermediary nowhere.
• A quotation attributed directly when it was taken from a summary.
• Citing a textbook's summary of a study, then discussing that study's method as if you had read it.

The rule that settles most cases

If you would be embarrassed for your marker to ask you a detailed question about the original — its sample, its method, its argument in chapter three — you have not read it, and you should be citing secondarily.

That test is uncomfortable and it is reliable. It also points at the fix: for the two or three sources your argument genuinely rests on, go and read them. For the rest, an honest secondary citation is not a flaw, it is a description of how research actually happens.

Translations are a special case

Reading Foucault in English is not a secondary citation. A published translation is an edition of the original work, cited as the original with the translator named, and you cite it directly because you read the actual text — someone else's rendering of it, but the whole text, not a report of it.

The distinction that matters is between reading a translation and reading someone's account of a work you cannot read. The first is direct citation. The second is secondary, and in disciplines with a substantial untranslated literature it is one of the four genuinely acceptable cases.

Where a term is contested in translation, name the original word in brackets on first use. Reviewers in philosophy, theology and comparative literature look for exactly this, because it shows you know the translation is an interpretation.

Cite the translation you used, with its own year, and give the original publication year where your style provides for it — APA uses the form (Foucault, 1975/1977), which tells a reader at a glance how far apart the two are.

Reference managers get this wrong

Zotero, Mendeley and EndNote have no field for 'cited in'. If you drag in the original record and let the manager format it, you will produce a direct citation of a work you never opened, and the software will do it silently.

The workaround is to store the intermediary as the record and write the secondary attribution into the text yourself, either by hand or through the suffix field that all three provide on the citation dialog.

Check every secondary citation in the final PDF rather than the working document. A manager that refreshes fields on export can quietly drop a manual suffix, and by then you have submitted.