Is "running" a gerund or a participial adjective?

1. An enlightening experiment

Google Books yields only 39 results, and instead asks me if I wanted to say “an enlightening experience”, and eagerly shows an impressive 10,000 results when I click on their suggestion. But for my purposes, I want the noun experiment

2. A running experiment

Dictionary.com suggests that running in 2. must be an adjective

adjective

18. prevalent, as a condition or state:
    running prices.
19. going or carried on continuously; sustained:
    a running commentary.
20. extending or repeated continuously:
    a running pattern.

We normally hear the plural form, running experiments, but the singular form is grammatical, and Google Books produces about 54 results.

Enlightening in "An enlightening experiment" is an adjective because I can use "very" to modify it:

A very enlightening experiment (YES)

but in sentence No.2, running cannot be modified by "very"

A very running experiment (NO)

It makes no sense, even though "running" appears to be an adjective. The comparative form,
“A more running experiment” is evidently wrong, but with enlightening, the comparative is acceptable:
“A more enlightening experiment”.

• Does that mean running in sentence No.2 is a gerund?
• How can I prove that running is a participial adjective or a gerund?

This question was inspired by this ELL answer

Solution 1:

tl;dr

• Despite running being in origin the -ING inflection of the verb to run, in your “a running experiment” example, it is no longer a verb and therefore ᴄᴀɴɴᴏᴛ be either a gerund or a participle either. A rule of thumb is that “No verb = No gerund–participle”.
• Here it’s almost certainly an adjective because it passes the Predicate Test, but it may under certain exotic readings be a noun instead.
• The OED calls these former -ING verbs participial adjectives and deverbal nouns respectively.

Your Question

You’ve asked whether in

• a running experiment

The word running is a “gerund” or a “participle” — or an adjective, and you propose a gradability test to see whether something is an adjective. This bothered you because as an adjective, running isn’t all that gradable.

Gradability is one test for adjectives, but not the only one. Another and perhaps stronger adjective test is the predicate test. This test distinguishes adjectival uses from nouns used attributively.

The Predicate Test for Adjectives

The way this one works is that given a noun phrase containing a pair of words “Y X”, where X is a noun but you aren’t sure whether the Y in front of it is an adjective, you look at the copula created by placing Y in the predicate positions and saying that “X is Y”. Then you decide whether that copula means the same thing as the original Y–X pair means. If so, then Y is (probably) an adjective. If not, then something else is going on here.

First, let’s show that the test works for other cases. When one says that “someone is in the dog house”, you know that with dog house you have a noun phrase here because it has a determiner (the definite article) and it is itself the object of the preposition in. But you don’t know whether in dog house, that first word dog is a noun or an adjective. The copula you need for the predicate test is therefore “house is dog”.

Does it still mean the same thing? Can you switch “dog house” to “house is dog”? Try it:

• someone is in the dog house
• someone is in the house (that) is (a) dog

No, you cannot, because in fact the house is not a dog at all! Rather, the house is for dogs! Dog is not an adjective. Rather, dog is a noun in attributive use. Further evidence that dog is a noun here is that you can restore the original sense by inserting a preposition (ᴘʀᴇᴘ) like for, by, with, from in your “Y is ᴘʀᴇᴘ X” copula. That means you know you have two nouns.

A running experiment

So, let’s try it. Create a “X is Y” copula:

1. running experiment
2. experiment is running

Does that mean the same thing? Does it make sense? If so, then running is an adjective.

I rather expect this to be the case you mean here. There is, however, another possible read, one which needs ᴘʀᴇᴘ to make it read correctly in the copula:

3. experiment is about running

Are you doing experiments on running? If that’s the sense you mean it in, then no, running is not an adjective there, but rather a noun.

Wait, I asked about participles versus gerunds!

So you did, so you did. Fine, but first let’s establish some ground rules to make sure we agree on what we’re talking about:

1. A gerund is a verb in its -ING inflection that heads a verb phrase used somewhere the grammar requires a noun phrase.
2. A participle is a verb in its -ING inflection that heads a verb phrase used somewhere the grammar requires a modifier phrase.

Yes, that’s a mouthful, which is why I wanted to make sure we understood the difference between nouns and adjectives first before addressing the question asked using the terms it asked about.

It’s such a mouthful that people often cut corners and pretend that gerunds and participles are nouns and adjectives. But they are not. Gerunds and participles are both verbs. They are neither nouns nor adjectives. The only difference is that the verb phrases they head are used in place of a noun phrase in the “gerund” case versus being used in place of a modifier phrase.

So a gerund phrase is an -ING verb phrase used substantively while a participle phrase is an -ING verb phrase used um “modificationally”. (SORRY!)

There isn’t a lot of difference there, and sometimes the example is simply too short to say one thing or the other. But what’s key is that you can use the predicate test here to winnow out the noun-like gerund-phrase case from the adjective-like participle-phrase case.

However, you’re going to have to apply some verb tests first to make sure that these are really verb phrases. If you can’t get any of the verb tests to pass but you can get noun or adjective tests to pass, then you do not have a verb. You just have a noun or an adjective.

The Object Complement Test for Verbs

One of the best verb tests we have the object complement test. Does the would-be verb take a direct object and perhaps an indirect object? Of course this only works for verbs that can ever take objects at all, meaning transitive verbs, but so be it.

Running can take an object, because you can run errands or races or drugs. It can even take an indirect object if you’re running someone an errand. Is this an experiment about running things?

If this were an experiment about running drugs, then sure. If you put your drugs object in front of the verb as is customary, then you have a drug-running experiment.

Notice though that I’ve needed that pesky preposition about there. The predicate test fails without it! So this verb phrase is not being “used as an adjective”. Rather, this verb phrase is being used substantively, being used in place of a noun phrase.

That leads us down the gerund fork not the participle fork.

An aside for noun tests

Understand that drug running is a compound noun, because running is a noun not a verb. You can tell by using noun tests on it. A noun has a particular slot in a noun phrase, so if you can use running there, then it is a noun. Determiners like his and adjectives like constant come before nouns, so try it:

• His constant drug running doomed him to an early death.

That proves that running is a noun. So does this:

• His constant running of drugs doomed him to an early death.

In fact, that’s still a noun without the drugs involved.

• His constant running doomed him to an early death.

So those runnings are nouns. That isn’t what we have here though. Probably.

Fine, fine: but which is it, gerund or participle?

I actually don’t think it’s either, because it’s not a verb here. There aren’t any objects, and really, you can’t put any there either, not in the normal way.

So I don’t believe that running is a verb in your example of a running experiment. According to the ground rules we established at the top of this section, it has to be a verb to be either of those two things. And since it is not a verb, it must be neither.

Here running passes the object-complement test for verbs (drugs being the object*, so here it is indeed a verb:

• He has been running drugs all his short life.

This is also a verb example, but different:

• What scared her the most was him constantly running drugs instead of making an honest living.

Both those are examples with running as verbs with verbal complements, which means that as verbs they’re so-called gerund–participles not nouns or adjectives, with no finer distinction necessary. It’s that they’re verbs that matters.

But you don’t have a verb in your example, only a former verb that’s now either a noun or an adjective.

So are my running experiments using running as an adjective or a noun?

If you can get it to pass the predicate test without having to add a preposition to connect two nouns, then you’re probably talking about an adjective here.

Which of these two works?

• the experiment is running
• the experiment is about running

It’s your experiment so you’ll have to tell me, but I’m going to guess that the first applies and the second fails.

That means that running is here an adjective.

Which, you’ll note, is what your dictionary entry duly reported in the first place.

Mind you, many dictionaries get this sort of thing wrong or gloss over it too much for a correct reading. But this one, I think, has the right of it.

What the OED says about these

The OED categorizes -ING words that are no longer verbs as nouns and adjectives, but it qualifies that in some cases calling them verbal nouns and participial adjectives. (BTW, you’ll find other sources referring to ex-gerunds as de-verbal nouns instead of as verbal ones. Same diff.)

And in point of fact, running may be either of those according to the OED. It has one entry for running labelled “ppl. adj.” but another entry for running labelled “vbl. n.”

There are lots of ex-verbs that can have both flavors:

• acting: an acting mayor (ppl. adj.) versus an acting school (vbl. n.)
• living: a living creature (ppl. adj.) versus a living space (vbl. n.)
• singing: a singing bird (ppl. adj.) versus a singing lesson (vbl. n.)

Again, these are all ex-verbs, so they cannot be gerunds or participles. What’s left behind once you’ve defrocked them of their verbal entourage is just an adjective or a noun, not a verb.

Ear training exercises

Here’s something that may help here, at least if you’re a native speaker. In speech, native speakers easily create and recognize the distinction of using an -ING (participial) adjective versus using using an -ING (verbal) noun there using stress. Except under emphasis, the adjective -ING word in front of the noun doesn’t take primary stress in the phrase, but the noun one does: a singing BIRD versus either of a SINGING lesson or a SINGING LESSON.

Solution 2:

In your examples, "enlightening" is best seen as an adjective and "running" as a VP comprising a gerund-participle form of the verb as head.

Taking "running" first: it fails the usual tests for adjectivehood: (a) it can’t be modified by "very" (* The very running man); (b) it can’t occur as complement to complex-intransitive verbs like "become" (* It became quite running); (c) it can’t occur a complement to complex-transitive verbs like "find" (* It found it quite running).

Other examples of VPs as modifiers are an approaching train, a sleeping child, some melting ice-cream.

"Enlightening" on the other hand passes the tests: It can be modified by "very", which can’t modify verbs (A very enlightening experiment); and it can occur as complement to both complex intransitive verbs and complex-transitive verbs: (It became quite enlightening; I found it quite enlightening). Other examples are a very entertaining show; a very fightening film.

That can only mean that "enlightening" when it occurs as an attributive noun modifier is an adjective, while "running" is a gerund-participial verb as head of a VP.

Solution 3:

1. I saw a running man. 2. There was a running experiment. 3. He had an extremely running (runny) nose.

In No. 1, the word running is definitely a present participle (verb). However, it is difficult to say it is a present participle in No. 2 and No. 3 as the verb run is used metaphorically and it can be changed to "on-going, continuous, etc." in No. 2 and "mucus-discharging" in No. 3. In other words, an experiment and a nose can't physically run and both of them are used metaphorically. The distinction doesn't seem to be very clear and whether it is a present participle or an adjective will depend on how you interpret the verb run in each example.

It's somewhat wrong to assume that all adjectives in English can take "very" as an adverb to emphasize their meaning and "more" to construct a comparative sentence. Some adjectives, especially those derived from -ing forms or past participle forms, don't take either. For example,

He is married.

The adjective married means "united in marriage". But the following sentences sound unidiomatic unless they are used metaphorically:

*He is very married. *He is more married than she (is).

The reason is the adjective married itself describes a state that can't be emphasized further or compared. In order for such an adjective to take either "very" or "more", it should get some help from another adjective or adverb, e.g., happily as in:

He is very happily married. He is more happily married than she (is).

The adjective running belongs to the same category of adjectives that can't take either "very" or "more". If you want to use one, it should take another word such as "long" or "short" as in

Very long (short)-running experiment. Longer (shorter)-running experiment than X.

Some grammar books call them "ungradable or non-gradable adjectives":

Some adjectives describe qualities that are completely present or completely absent. They do not occur in comparative and superlative forms, and cannot be used with adverbs such as very or extremely, because we don’t usually imagine degrees of more or less of the quality being described. They are referred to as non-gradable adjectives. Non-gradable adjectives do sometimes occur with non-grading adverbs such as completely which emphasize the extent of the quality.

(emphasis mine)

Some ungradable adjectives described in this link can take "very" and "more". There seems to be no hard-and-fast rule and it will entirely depend on whether an adjective sounds idiomatic or not with "very" and "more".

Solution 4:

There is much good discussion in the other answers. This is an historical note.

In Aspects of the Theory of Syntax (1964), Chomsky considered the phrase "the sleeping child", and argued that it is transformationally derived from ?"the child sleeping" by Modifier-Preposing, and in turn that comes from the relative clause construction "the child who is sleeping" by Relative-Clause-Reduction (or WHIZ). I am giving here a rather free paraphrase of Chomsky's argument.

He gave as evidence for his analysis, the ungrammaticality of *"the child who very sleeps" and *"the child who is very sleeping", which is obviously due to the fact that "very" can't modify verbs. From this constraint, and from the transformational derivation above, it follows that *"the very sleeping child" should be ungrammatical. Thus transformational grammar (TG) explains why "very" cannot modify the modifier "sleeping". In traditional grammar, this fact appears mysterious and would require an ad hoc constraint.

Is this a fair argument by Chomsky? Yes, but only in (what we would now call) classical TG. In the theory described by Chomsky, subcategorization constraints based on lexical categories (the parts of speech) have meaning only for deep structures, before any transformations apply. It is not necessary to appeal to a category of "participle-modifiers" to explain special properties of the "sleeping" modifier, and in fact it isn't even possible. Lexical constraints like the one under discussion are relevant only to deep structure, and in deep structure, there is no modifier "sleeping" -- only the progressive aspect of the verb "sleep", which occurs within a relative clause.

I think this is an important difference to know about between traditional grammar and TG. Above, I use the phrase "classical TG" sometimes, because not all contemporary forms of TG are like the original version of TG as described by Chomsky in Aspects. McCawley's version of TG, for instance, is deliberately different in this respect from the classical TG of 1964. As for CGEL, it seems to be a hodge-podge; I doubt the authors know what syntactic theory they are using.