gerunds - Is "programming" not a noun?

Recently, I was told that the word "programming" in the phrase "programming thoughts" is a verb in the gerund-participle form and that the term "gerund" by itself is obsolete in modern grammar. I was confused by this, because to me the construction seems exactly the same as in, say, "Shower Thoughts": a noun being modified by an attributive noun. I understand that it's the same form as the present participle. But it seems to act exactly as a noun phrase: for example, it can be modified by adjectives like "good" or "bad." I know that some -ing words can be analyzed as deverbal nouns; this is described here: what is the difference between a “deverbal noun” and a “verbal noun”?

However, the tests given seem to mainly apply to count nouns; normally we can't pluralize mass nouns, and they can't take the indefinite article a/an. The definite article can be used with programming; for example, in the phrase "the programming of computers."

So why can't it be considered an indefinite deverbal noun? Is that even the right terminology? Is there any terminology in modern theories of syntax by which one can distinguish the "programming" in "programming class" (a class about programming) from the "programming" in "programming grandparent" (a grandparent who programs), or are these not considered to be distinct constructions?

I've read a Language Log article about the concept of gerund-partiple and how it covers some of the things traditionally labelled as "gerund" (Gerunds vs. participles) but the examples of the "gerund" here seem to be more verby than "programming":

Destroying the files was a serious mistake.
I regret destroying the
files.

Also, in a comment, Mark Liberman says

Huddleston and Pullum argue, in addition, that the traditional
functional distinction between gerunds and present participles is not
a coherent one, and should be abandoned. That's not the same as
arguing that all -ing forms are gerund-participles — this is clearly
false — or there are no functional distinctions among uses of
gerund-participles — H & P retain or propose several, just not

anything that's closely congruent with the traditional split.]

What are the -ing forms that are "clearly" not gerund participles? And what are the names Huddleston and Pullum use for the functional distinctions?

This Wiktionary discussion suggests that H & P distinguish "gerundial nouns" from "gerund-participles" using the following criteria:

• Complementation: Among other things, gerundial nouns take an "of" -prepositional phrase as complement ("the singing of the song").



• Modification: Gerundial nouns take an adjective as modifier, while participles take an adverb. ("Her splendid singing of the song left
  them transfixed.")


• Determiners: Only nouns can take the definite article ("the singing of the song").


• Plurals: Only nouns can take the plural (however, they don't always do so).

By these criteria, it seems to me that "programming" is a gerundial noun, not a gerund-participle. And assuming the term is compositional, I'd think a "gerundial noun" can be called a noun. But I'm not sure I'm getting this right.

I guess the same question applies to words like "cooking" and "writing."

---

(I know that many dictionaries list "programming" as a noun, but dictionaries can't be trusted to get parts of speech technically correct.)

Answer

I never used the terms "gerund" and "participle" when I was learning syntax, nor when I was teaching it, so I think we could easily do without those terms. However, on the other hand, I don't see a problem with the terminology, provided that one is careful.

For the English forms derived from verbs by adding "-ing", there are:

A. nouns. These take articles, can be modified by adjectives (like any ordinary noun), but do not take direct objects (though sometimes logical objects can be expressed with "of"). They cannot be modified by adverbs, just as other nouns cannot.

B. gerunds, which are verbs used as nouns. These are the heads of nominalized sentences (so that is why they are said to be used as nouns), but like other verbs, do not take articles, cannot be modified by adjectives, can be modified by adverbs, and if transitive, can take direct objects (unlike true nouns).

C. participles, which are verbs used as adjectives. These are either like predicate adjectives, used for the English progressive construction, or noun modifiers. Unlike true adjectives, they cannot be modified by "very". When they are noun modifiers, they can be thought of as coming from reduced relative clauses, and when they are alone, without a direct object or other complement, they will be preposed to the noun they are predicated of. [Chomsky gave the example "the (*very) sleeping child".] If you think of a noun modifier as an adjective, then these are derived adjectives.

D. adjectives. These can usually be modified by "very", but cannot take direct objects. Compare the true adjective "annoying" in "He is very annoying to me", with modifier "very" but without a direct object, and the participle "annoying" in "He is annoying me", which cannot have "very" and does have a direct object.