Traditionally, a clause is defined as consisting of a subject and predicate. In Oxford Dictionary, it is defined as:

A unit of grammatical organization next below the sentence in rank and in traditional grammar said to consist of a subject and predicate.

But recent linguistics says you don't need a subject to qualify as a clause. For example, an imperative clause normally can get away with no subject:

Come on in.

Except for the imperatives, subjectless clauses are essentially non-finite clauses having an infinitive or participle form of a verb.

He wanted [to change his name]. [infinitive]

She was [locking the door]. [participle]

His father got [charged with manslaughter]. [participle]

These bracketed portions all have a verb as its head. That is, they're all verb phrases, which still holds water even in recent linguistics and outside traditional grammar.

Now, what's the merit of reclassifying these verb phrases as non-finite clauses? To me, it merely seems to make the already confusing phrase-clause distinction even more confusing.'

  • 4
    Wouldn’t this be a better fit for Linguistics? The same considerations apply to many other languages, not just English, and I think you’d be more likely to get answers with a thorough theoretical base over there as well. – Janus Bahs Jacquet Feb 27 '18 at 8:09
  • 1
    Just flag for moderator attention and ask them to migrate it. – Janus Bahs Jacquet Feb 27 '18 at 8:12
  • 1
    I'll close-vote unless you remove the grouse 'Now, what's the merit of calling these verb phrases as non-finite clauses other than making the already confusing whole phrase-clause distinction even more confusing? ' // 'Now, what's the merit of reclassifying these verb phrases as non-finite clauses? To me, it merely seems to make the already confusing phrase-clause distinction even more confusing.' de-grouses and retains valid content. //// You've not mentioned verbless clauses. – Edwin Ashworth Feb 27 '18 at 10:28
  • 2
    I'm just saying I would like to know the answer to your Q and, for me, the question is on topic. You have to make that call, it's your question, not anyone else's. You know both sites well enough to make an informed decision. – Mari-Lou A Feb 27 '18 at 10:48
  • 1
    ' ... what's the merit of calling these verb phrases as non-finite clauses other than making the already confusing whole phrase-clause distinction even more confusing?' is unwarranted sarcasm (or perhaps merely poor writing, I'll concede): how can 'making the already confusing whole phrase-clause distinction even more confusing' be included as a 'merit'? I've edited. – Edwin Ashworth Feb 27 '18 at 11:05

Well, this is a question about grammatical terminology, not about grammar, or English grammar, as I understand it. I've looked at a bit of the prior discussion here, comments and answers, but I'm not replying or referring to any of that here. I'm afraid I'm also not replying or referring to @JK2's concerns, either. This post is strictly intended as an answer to the title question, more or less as stated.

What is the merit of calling a verb phrase a clause?

But let's leave out the "merit" part; that's a preachy word in my English -- substitute "purpose", since this is something linguists do on purpose. And let's also not limit the answer just to verb phrases and clauses; if possible, a general principle -- or at least some well-plumbed rules of thumb -- would be welcome.

In the case of the verb phrase and the clause -- the presenting problem, as it were -- the question to ask, once you're able to identify verb phrases reliably (which takes some practice), is

  • What is a verb phrase, anyway?

There are a lot of things, and kinds of things, that are identifiable as verb phrases. A vast number of them. How do we learn to recognize them? Well, we tend to find them in clauses. Clauses are not really complete without a verb phrase, especially in English. In Even Cowgirls Get The Blues, Tom Robbins makes this point with the following statement:

  • This sentence no verb.

Probably the simplest and most useful description of a verb phrase is

  • A verb phrase is a clause that's missing its subject.

In other words, if you have a verb phrase, all you need to do is supply a subject, and you have a clause. And supplying a subject is never a problem -- we're good at that. That's why people continue to dangle participles and make dumb jokes -- they know their addressees will figure out what they want them to.

So here's the situation that puts a syntactician in. (One knows, of course, but doesn't particularly care, how The Oxford Dictionary defines phrase or clause; that's up to the syntactician to tell the dictionary, not the other way around.) The purpose of grammatical terminology is to make things easier to understand and remember. If verb phrases are associated with clauses, and one can always produce a clause with an appropriate subject for any verb phrase, why not say that verb phrases are just clauses with pieces missing? What would you do if you found a bird wing on the ground? You'd look around for the rest of the bird.

And so certain types of syntacticians (of which I am one) will tell you that every verb phrase in a sentence defines its own clause. Most of these clauses lose pieces because the pieces are predictable (by rule -- discovering these is what syntacticians do) and therefore recoverable (or else unimportant and therefore ignorable). This is basically what McCawley's grammar of English describes.

As it turns out, there are a lot of regularities. For instance, tensed English clauses have to have subject NPs present, to the point where dummy NP subjects like it and there are required, not optional, in many cases. On the other hand, untensed English clauses like infinitives, gerunds, and participles are frequently missing subjects and other chunks.

  • Bill wants [for Bill] to win = Bill wants to win.
  • Bill wants [for] Bill to win = Bill wants himself to win.
  • Bill wants (for) Mike to win = Bill wants Mike to win. (optional for)

These are handy and pop up all over the place; think about the purpose infinitive, the relative infinitive, and all the gerunds and participles people use for all kinds of purposes. They all represent longer clauses, and these clauses can be exhibited as needed. For a syntactician, that means an easy way to relate and categorize a lot of different syntactic phenomena, all under one rubric. That's irrestistible.

And that's the title McCawley chose for his grammar of English:

One could go on in this vein for quite a while, but I'll content myself with the observation that this is the same purpose, and the same kind of thinking, that allows clauses to function as noun phrases, and as adverbs, and as modifiers in a noun phrase. You use the tools to hand.

  1. Grammar rules of some languages even traditionaly had to acknowledge one-terme clauses. For example Czech (and other Slavic) sentence "It is raining" contains only the verb, and supplying subject would look as strange as "Something is raining" in English. Oxford dictionary today should acknoledge other languuages have clauses, too.

  2. Linguistic research today studies the language as it is produced by people, not creating ideal language rules that nobody precisely follows. So anything people say, newspaper print needs to be classified. "Well!" is a perfect part of the language, as well any heading in newspaper or book. And I guess that they get classiffied as clauses (or sentences). "10, 9, 8, 7, 6, 5, 4, 3, 2, 1, go!" sounds to me as eleven clauses.

There is also preasure to have reasonable technical terminology when audio or text are collected in corpuses, and everything has to be segmented and classified. Postal address is a part of language, as well as "KarlG Feb 27 '18 at 6:37" from this page, or a chess game record. With a project of automatic translator in mind, or information retrieval engine, one should rather include everything.

  1. The linguists recognise about three levels of language, and when you go from the deep (near semantic) level to the surface level (what you say or write) there are several ways how a term may get lost (because known from the context, by analogy, because of language usage...). Any of them apply, generally, also to subject and preditate, and can make them disappear. However, there are some specific properties of English like that the subject is forced to appear on the surface level, and it is eventually "created" if it not at hand. ("It is five o'clock." "What is five o'clock?" "The time!" "So, why don't you say <The time is five o'clock>?".) This makes non-traditional clauses in English may-be infrequent but not impossible. So what is full subject-predicate on the deep level, may miss something on the surface level.

Sorry for my poor English, and my poor linguistics, but I hope my answer might help a bit. If you need better answer, maybe I convinced you to ask in the Linguistics department.


The merit behind this non-finite redefinition is to better acknowledge that verb usage can apply in those situations with context, and those without, of which both work by modern linguistics and must require an adjoining subject target to form a complete conclusion. Non-finite meaning without an end or conclusion, as without the adjoining subject target, there is none.


  1. One which must contain a contextual definition of the subject target in a prior sentence or phrase. (participle, subject-less, adjoining phrase)
  2. One which must contain a contextual definition of the subject target in the current sentence or phrase. (participle, subject-less, compacted adjoining phrase)

Both of these scenarios are non-finite situations requiring additional phrases to offer a conclusion by the use of non-finite clauses.

Examples from yours meeting the subject-less clause rule of verb usage:

She was [locking the door]. [participle]

This fits into a subject-less clause adjacent to a contextual subject inquiry.

("What is [she] doing?") he asked. [contextual target] (inquiry)

("[Locking] the door.") [subject-less verb] (Subject-less clause)

It requires the first phrase prior to the non-finite subject-less verb usage to offer a conclusion.

His father got [charged] with manslaughter. [participle]

This fits into a non-committal subject-less clause adjacent to a contextual subject inquiry, with the addition of the ellipsis '...', implying non-committal or detached demeanor in the reply.

("What happened to [his father]?") [contextual target] (inquiry)

("...[charged] with manslaughter.") [subject-less verb] (non-committal reply)

It requires the first phrase prior to the non-finite subject-less verb usage to offer a conclusion, and provides emotional implications with the addition of the ellipsis.

So in other words, non-finite is describing the fact a subject-less clause must contain an adjoining phrase to become a finite body of words offering a conclusion. It is indeed difficult to comprehend as it deals with bodies of words, not entirely one sentence in many cases, and requires context.

Simply remembering "my non-finite verb-clause needs a finite conclusion to work" is a general mental rule that works.

  • Sorry, but I don't know what the heck you're talking about. What do you even mean by 'conclusion' or 'end'? – JK2 Mar 19 '18 at 1:03

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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