Et al. done right

"Et al." looks like a convenience and is actually a rule. How many authors it may stand in for, when it may first appear, and whether the reference list is allowed to use it at all are all decided by your referencing style — and the styles disagree with each other in ways that produce real errors rather than stylistic ones.

It is also the single most information-destroying thing you can do to a citation, which is why it is worth getting right.

What it abbreviates, by style

The rules differ on two axes: how many authors trigger the abbreviation, and how much of the list survives it.

APA 7
  in text        three or more authors → first author
                 plus "et al." from the very first citation
  reference list up to twenty authors listed in full

Vancouver / NLM
  in text        a number — authorship is not shown at all
  reference list up to six authors, then "et al."

MLA 9
  in text        three or more authors → first plus "et al."
  reference list first author, then "et al."

Harvard
  varies by institution — commonly three or four authors
  in text; check your own handbook, because there is no
  single Harvard

Two things in that table cause most of the trouble. APA abbreviates in the text and expands in the list, so the same work looks different in the two places by design. And Vancouver hides authorship entirely in the text, which means an "et al." problem in a Vancouver thesis lives exclusively in the reference list where nobody re-reads it.

The rule people most often get wrong is APA's from the very first citation. Earlier editions asked for the full author list on first mention and the abbreviation afterwards, and a great deal of published writing still does it that way. If your handbook says APA 7, the first citation of a three-author work is already (Okafor et al., 2021) — there is no full-list first appearance.

The same work, two forms, one document

Here is the situation that a reader handles without noticing and a checker cannot.

Chapter 2   (Okafor, Lindqvist, & Duarte, 2021)
Chapter 5   (Okafor et al., 2021)

A human reads these as the same work and is right.

A matcher sees:
  citation A   three named authors, year 2021
  citation B   one named author, year 2021, abbreviated

Nothing in B says which work it abbreviates. It says
only that the lead author is Okafor and the year is
2021 — which is a description, not an identity.

This is normal and correct in several styles: a thesis written under an older APA edition, or under a house style that asks for the full list on first mention, will contain both forms of the same reference legitimately. The document is not wrong. But the abbreviated form has thrown away the two co-authors that distinguished it, and they are the information a comparison needs.

Et al. is a compression, and like every compression it is lossy. For a reader who can go and look, the loss costs nothing. For anything comparing citations against a list, it costs exactly the part that was doing the identifying.

When it becomes genuinely ambiguous

Most abbreviated citations can be expanded by looking at the rest of the document — find the work whose author list starts with the named prefix, and that is what the "et al." stands for. It works because the abbreviated form is a prefix: the names it does show are the first names of the real list, in order.

It stops working when more than one work fits.

Same first author, same year, in one thesis:

  Okafor, T., Lindqvist, M., & Duarte, P. (2021)
    Medication adherence in community settings

  Okafor, T., Chen, R., & Bello, A. (2021)
    Staffing variance in regional centres

And somewhere in chapter 6:

  (Okafor et al., 2021)

The prefix "Okafor" fits BOTH. Nothing in the citation
distinguishes them, and no amount of care by the person
checking will recover what the abbreviation removed.

When that happens, the honest answer is that it is ambiguous — and the rule we apply is that ambiguity is handed on rather than resolved. Two candidates means the pairing is shown to a person with both options, not decided by picking the more likely one. Guessing here would produce a confident, wrong attribution that reads exactly like a correct one, which is the failure mode surname-and-year matching produces at scale.

This is also, incidentally, what styles are trying to prevent when they ask for a year suffix. If you have two Okafor 2021 works, APA wants 2021a and 2021b — and those suffixes are not decoration, they are the identity the abbreviation removed, put back.

"Ambiguous" is not the same as "wrong"

Worth saying plainly, because a report that flags several of your citations as ambiguous reads alarmingly and usually means nothing is wrong at all.

An ambiguous abbreviation is a style artefact. You followed the rule — your style told you to write "Okafor et al." and you did — and the rule happens to be lossy in the specific case where you cite two works by the same lead author in the same year. The citation is correctly formed. The document simply does not contain enough information to say which of two works it means, and neither would any reader.

AMBIGUOUS       two works in your list fit this
                abbreviation
                → add a year suffix. Ten seconds.

NO ENTRY        the abbreviation fits nothing in your
                bibliography
                → the reference is missing. Add it.

UNRESOLVABLE    the work the entry describes cannot be
                found to exist anywhere
                → a different order of problem

The first two are bookkeeping and the fix is obvious once you can see it. The third is not about "et al." at all and is the only one whose consequences reach beyond your own submission — a reference to work that was never written is academic misconduct wherever it is found, however it arrived in a list. It is worth knowing which of the three you are looking at before deciding how much of your afternoon it deserves.

And a note on year suffixes for the ambiguous case, since it is the fix people put off: adding 2021a and 2021b is not merely a formatting nicety demanded by a handbook. It restores the identity the abbreviation removed. Once the suffix is there, "Okafor et al., 2021a" is unambiguous forever, in your document and in anything anyone builds on top of it.

Four rules that avoid all of this

One: settle your style before you write, not before you submit. Every rule above depends on which style you are in, and retrofitting "et al." conventions across four years of chapters is a genuinely miserable week. If you are not sure which style your document actually uses — as opposed to which one your handbook names — that question is worth answering first, because everything downstream depends on the answer.

Two: add the year suffix the moment you have two works by the same lead author in the same year. Not later. The suffix is cheap when you add it and expensive when you retrofit it, because retrofitting means finding every affected citation and deciding which of the two it meant — years after you knew.

Three: never abbreviate in the reference list unless your style tells you to. APA lists up to twenty authors in full; Vancouver stops at six. An entry that reads "Okafor, T., et al." in an APA reference list is missing information no reader can recover, and it is the one place the full list was supposed to survive.

Four: be consistent within a document, even where the style permits latitude. Mixing the full-list-first-mention convention with the abbreviate-always convention is not an error in most handbooks, and it doubles the number of forms every work can take.

The one that is not your fault

A last category worth naming, because people assume it is a mistake of theirs. Some works genuinely have enough authors that no style shows them all in text — large collaborations, consortium papers, clinical trials with forty contributors. A citation to one of those is abbreviated by necessity rather than convention, and the corresponding reference entry may itself be truncated by the publisher.

There is nothing to fix there. What is worth doing is making sure the entry carries an identifier, because when the author list cannot identify a work, the DOI is the only thing that can — and an identifier is checkable in a way an abbreviated author list never is.