The main difficulty results directly from your terminology, which as far as I can tell, is an unusual* one that is leading to paradoxes and misunderstanding. You appear to be using the term “gerund” in a purely inflection sense, not one related to either of part of speech or grammatical function.
- (It's certainly not unheard of to use it your way; precedent exists. And we do need a word for this. But it's so much more common to use it for one of the other two things that people wouldn't understand you meant it not as they use it but only as a name for the inflection.)
I imagine that this is what’s getting you into trouble because using morphology alone in English is seldom fruitful for higher level analyses. This is because invisible “zero-derivation” transformation to another part of speech is so common in our language.
Barring only defective verbs such as modals, you can always take the base form of any English verb and apply a completely “regular” type of inflectional morphology to it to produce an ‑ING inflection of that verb. Inflections of a word do not change its part of speech in and of themselves.
So if you start with a base form like speak, you can inflect that verb in just a few different ways in English, producing things like speaks, spoke, spoken, speaking. The first two inflections are always used as verbs, but the third and fourth inflections can also be used as other things besides verbs alone, mostly nouns or adjectives like givens and interesting but in rare cases other things like prepositions.
And this is precisely where terminology choice can become hazardous. Because if you're not careful, you will confuse parts of speech with grammatical roles. This is more perilous with the fourth form, the ‑ING one, than the third form that we use when there are auxiliary verbs involved.
Applying a few extremely basic syntactic tests will almost always show you which part of speech between noun, verb, or adjective a given instance of an ‑ING form really is.
If your ‑ING word can be modified by an intensifier like very or rather or fairly, then it can only be an adjective, not a noun or a verb. You can do that with phrases like very interesting, which shows you that here your ‑ING word must be an adjective. (There are a few uncommon cases where an ‑ING might be argued to be an adverbial acting as a “sentence adverb”, but we won’t worry about those rare birds here. Adverbs are still modifiers, just as adjectives are. But even then it could still be a verb clause acting in a modifier role that's adverbial, which wouldn't make it an adverb.)
If your ‑ING word started out life as a transitive verb and you can use whatever core arguments that verb normally takes, then it is still a verb. Being able to use arguments is one of the clearest signals we have that something is a verb. So in speaking truth to power, you know that this is a verb.
Another possibility you can try with these is to use an -ly adverb, but this alone is not dispositive because most of those can also modify adjectives, not just verbs. That's why we chose very for our adjective test, not some adverb. In calling loudly, you know that ‑ING word is a verb not a noun or an adjective.
Lastly, try applying the customary components of a noun phrase to it, and if those work, then it is only a noun, not a verb or an adjective. The easiest of these is to use nouns or adjectives attributively, so right in front of the ‑ING word. With bone healing you use an attributive noun bone, while with acceptable calling you use an attributive adjective acceptable. Those both show that this time those ‑ING words that started out life as verbs once upon a time must now be nouns.
Another way to show that they’re nouns is to see if they can be inflected into the regular plural. Only nouns do that, so if so, you can be certain you have an ‑ING word that’s really a noun, not a verb or an adjective. Think of earnings, for example. That’s a noun.
Still another way to detect nouns is to see whether you can use a prepositional phrase after the ‑ING word. If the original was transitive, see whether you can use of plus the object. That’s why it’s a noun in the herding of goats but just a verb in the activity of herding goats. (But if those goats were special goats bred specifically for herding, then the stress would shift to the other word, and herding would be a noun again.)
That’s really all there is to it.
You will perhaps notice that I haven’t used the word gerund in any of this. That’s because gerund is not a part of speech in English. It happens when you use an ‑ING clause (which being a clause is always going to have an actual verb at its heart, never a former verb) in a grammatical role fulfilled by a noun phrase, such as being the subject or object of another verb or the object a preposition. All these are grammatical roles; they are not parts of speech.
So your first example cannot be using an ‑ING word as a gerund because it is being modified by the noun bone. Therefore it is just a noun, and that noun phrase is the object of a preposition. But it is not a gerund.
Your other cases are using ‑ING words in modifier roles, not substantive roles, so they cannot be gerunds either. You are welcome to think of them as reduced relative clauses under whiz-deletion but you do not need to do that.
Your title asked about “noun + gerund”. None of your cases has that in it, but those can occur. They just aren't very common. That’s because now your -ING word needs that noun in front of it to be the subject of a non-finite verb clause. Here’s one such:
- People calling me after midnight really drives me crazy.
The subject of calling is people, while the subject of drives is the entire non-finite clause people calling me after midnight. It’s the number of the second verb that tells you what is meant here. Because it is singular, you know that the subject must be the entire clause, as clauses are always singular.
Your alternative proposed version is also valid, but means something completely different.
- People calling me after midnight really drive me crazy.
Now the verb drive is singular, and the only subject candidate that is plural is people. Now it is the people who are driving you crazy. In my version with the gerund clause, it is the calling itself which is driving you crazy, not the people doing it.
All these are -ING words used as gerunds. That means they are verb clauses in grammatical roles where only noun phrases work for, like subjects and objects. In all but one these verb clauses are the subject, and so their own verbs must be in the singular:
Healing bones takes forever.
Watching birds makes me happy.
I hate washing dishes.
Watching birds make me happy is my wife’s greatest joy.
(I deliberately chose transitive verbs with explicit objects every time so you could immediately see that the -ING words were all certain to be verbs there, never nouns or adjectives.)
The last one has three clauses instead of two like the others. Just count the verbs and you've counted the clauses.
Your own first example has only one clause because it has only one verb, is. None of the other words there is a verb, so that’s all the clauses it has.
I recommend that you stick to parts of speech like noun, verb, and adjective on the one hand and to grammatical roles like subject, predicate, and object on the other hand. (Yes, that still leaves a hole for what to call each inflectional form. Various choices exist for those, too.)
I know you want to say gerund or participle, but these are quite tricky to use in ways that will never confuse anyone because of the confusion over whether they refer to inflectional forms, to parts of speech, or to grammatical roles. Plus they aren’t strictly necessary. They’re also a bit old-fashioned, as modern analysis often avoids them altogether or fuses them into an unwieldy double⸗barrelled form, gerund-participle or even worse, gerundial-participial. No thanks.
I realize that saying an -ING word doesn’t look very formal and Latinate and all, but it better serves us in English. Leave Latin to Latin.
This question comes up many times a month here, often more than once in any given week. And I answer it every year or two, too. It seems unstoppable, and closing as the duplicate it really is at its heart never seems to do any good either. I have no solution to this problem, particularly as most reference materials for beginners treats this matter too simplistically to be of any practical use. Worse, almost all of them are actually flat-out wrong due directly to their wild oversimplifications and misrepresentations.