A lot of people seem to understand "matrix clause" as a synonym for "main clause". For instance, a comment I just chanced upon on a language SE site states:

It's a synonym for main clause (or independent clause) as opposed to subordinate clause (or dependent clause, or indeed embedded clause).

Even some Wikipedia articles seem to echo this:

Main clauses (matrix clauses, independent clauses) are those that can stand alone as a sentence.

That isn't exactly the way I understand the concept. I am under the impression that a matrix clause is a clause with an embedded clause, so it doesn't have to be a main clause, and on the other hand a main clause is not necessarily a matrix clause. Too bad matrix clause doesn't have its own Wikipedia page which would've settled this. But even the name matrix clause is an obvious clue as to its properties. As this discussion demonstrates, matrix comes from "mother" or "a female animal that gives birth to others", pointing to the fact that it is a clause with child clauses. Online sources1 also confirm my understanding:

A matrix clause is a clause that contains another clause. Thus, the main clause in (37), the professor told the students, is a matrix clause since it contains another clause... So the situation described in the embedded clause is contained by, and functions as an element of, the situation described by the matrix clause.

A matrix clause is often a main clause . . ., but it need not be: it can itself be a subordinate clause.

But it bothers me that my understanding and this definition run counter to that Wiki article. What exactly is a "matrix clause"?

1Books cited in this source:

  • Martin J. Endley, Linguistic Perspectives on English Grammar. Information Age, 2010
  • R.L. Trask, Dictionary of English Grammar. Penguin, 2000
  • 2
    A matrix clause is defined as a clause within which a subordinate clause is embedded. In, for example, I think [she said she was ill]", the bracketed clause is the matrix clause in which "she was ill" is embedded. The uppermost clause is often referred to as the main clause.
    – BillJ
    Jan 28, 2021 at 9:48
  • 1
    @BillJ My thoughts exactly. So I think it is fair to say a matrix clause doesn't have to be a main clause, and vice versa, right?
    – Eddie Kal
    Jan 28, 2021 at 9:50
  • 1
    Yes: When we define 'main clause' we are talking of one that is not embedded as a dependent within some larger clause. In the example above, I would say that the main clause is the whole sentence.
    – BillJ
    Jan 28, 2021 at 9:56
  • 1
    @BillJ This would make a definitive answer. But I know you sometimes seem to prefer to provide helpful information in the comments. Either way, thank you.
    – Eddie Kal
    Jan 28, 2021 at 9:59
  • 1
    This is the principle of the Derivational Cycle. Cyclic rules like Passive or Extraposition can be applied to each successive nested subordinate clause, starting at the bottom and proceeding up the tree to the top S. The term matrix clause refers, on each cycle but the last, to the S above the current cycle. In a simple sentence, there's no need for it, but it can function as a sloppy synonym for main clause, since there's no subordinate clause.
    – jlawler
    Jan 28, 2021 at 14:50

1 Answer 1


The Cambridge Grammar of the English Language (Huddleston & Pullum 2002) states:

The clause in which a subordinate clause is embedded is called the matrix clause -– in [1vb], for example, subordinate that Liz was ill is embedded within the matrix clause He said that Liz was ill. Subordination is recursive, i.e. repeatable, so that one matrix clause may be embedded within a larger one, as in I think he said that Liz was ill. (p. 47. Bold and italics as in the original)

So (1) below is a main clause:

  1. Liz was ill.

It isn't a matrix clause, however, because it contains no subordinate clause.

In (2) below, the whole sentence is a main clause and also a matrix clause and, as described above, that Liz was ill is a subordinate clause.

  1. He said that Liz was ill.

In (3) below, the whole sentence is a main cause and a matrix clause:

  1. I think he said that Liz was ill.

The sentence contains a subordinate clause, he said that Liz was ill. This subordinate clause has another subordinate clause embedded within it, that Liz was ill. The clause he said that Liz was ill is a matrix clause in relation to the smaller clause that Liz was ill.

We can see therefore that the term matrix clause is in some sense deictic. A clause is a matrix clause specifically in relation to the clause it contains. We could think of it as a 'container'.

  • 3
    Yes, and I think it makes sense to reserve the term 'main clause' for the overall uppermost clause, which here is the sentence as a whole, since it's the only clause that is not embedded as a dependent within some larger clause.
    – BillJ
    Jan 28, 2021 at 15:36
  • @BillJ I wonder… what would you call a non-embedded clause that begins with a subordinator, i.e., a typical subordinate clause used on its own? It feels intuitively wrong to call it a main clause, but it also feels wrong to call it a subordinate clause, since it’s not embedded as a subordinate entity within anything (except on a more abstract level in the larger discourse context). An un- or desubordinated clause, perhaps? May 22, 2022 at 10:32
  • @JanusBahsJacquet “Fragment” is the popular term, but it has its own pitfalls. May 22, 2022 at 14:14
  • @Araucaria-him Yeah, I can see how that would be pitfally. It not only implies but outright claims that it’s a part broken off of something, presumably (I’m guessing) that it’s a subordinate clause ‘broken off’ from its matrix clause, which isn’t necessarily the case. May 22, 2022 at 14:24
  • @JanusBahsJacquet Yes, exactly so. May 22, 2022 at 21:08

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.