Derivation of Nouns from Verbs in English

[Terra]

English seems to form nouns from verbs according to two different patterns:
1) verb -> tool (the noun means "that which x-es") (Perhaps this category is really verbs formed from nouns.)
Examples:
- hammer
- pin
- rake
- whip
- chain
- plough

2) verb -> action, result (the noun means both "the action of x-ing", and "the result of x-ing")
- cut
- break
- fold
- scratch
- crack
- furrow

Some verbs, seem to be deficient though, and the nouns they yield cannot be used for results, perhaps because the results are too ephemeral.
- punch
- kick
- smash
- crush
- crash?
- shake

Other verbs confuse me even more:
- thunder
- feed

Their respective derived noun of each is not a tool, action, nor result.

What do you think of my analysis? How would you classify the last couple of verbs? Is not knowing what pattern to follow a common stumbling block for L2-learners? (Note how "plough" and "furrow" mean nearly the same thing in their verb forms, but not so in their noun forms.) Is one pattern more common in languages than the other?

[linguoboy]

English seems to form nouns from verbs according to two different patterns:
1) verb -> tool (the noun means "that which x-es") (Perhaps this category is really verbs formed from nouns.)
Examples:
- hammer
- pin
- rake
- whip
- chain
- plough

A good dictionary will tell you which usage was attested first, nominal or verbal. According to my sources, all of these words were first found as nouns.

2) verb -> action, result (the noun means both "the action of x-ing", and "the result of x-ing")
- cut
- break
- fold
- scratch
- crack
- furrow

Unlike the others, furrow is attested first as a noun.

Some verbs, seem to be deficient though, and the nouns they yield cannot be used for results, perhaps because the results are too ephemeral.
- punch
- kick
- smash
- crush
- crash?
- shake

How are these verbs "deficient"? I cut something; the result is a cut in it. I kick something; the result is a kick to it. I'm not seeing the distinction between these and the previous batch that you do, not even with crash. ("Something crashed to the floor upstairs. I know; I heard the crash.")

Other verbs confuse me even more:
- thunder
- feed

Their respective derived noun of each is not a tool, action, nor result.

Again, thunder was a noun first. And I'm not seeing a difference between, say, thunder and crash.

What do you think of my analysis?

I think it leaves something to be desired.

[clawgrip]

Some of the category 1 words started out as nouns from which verbs were derived (such as hammer, rake) while others started out as verbs (such as whip, maybe chain). I think referring to the nouns here as tools is too specific, and more likely a subset of a larger category. I would describe the nouns in this category as simply being items that are closely associated with the action, and/or items frequently used to perform the action.

The verbs in (3) I think are just a subset of 2. For the most part, don't really leave any result that is clearly a product of the verb in question, so the "result" so to speak is the action at moment it occurs, which is the only time the action is clearly identifiable.

Thunder could be either 1 or (3), i.e. a noun closely associated with the verb, or a direct result of the verb.

'Feed' fits in category 1, because the noun is closely associated with and frequently used to perform the action in question.

[Terra]

How are these verbs "deficient"?

I'm using "deficient" to mean that they cannot by used as a result, but can as an action, compared to the second group which can be used as both a result and an action.

How are these verbs "deficient"? I cut something; the result is a cut in it. I kick something; the result is a kick to it. I'm not seeing the distinction between these and the previous batch that you do, not even with crash. ("Something crashed to the floor upstairs. I know; I heard the crash.")

I can point to my wounded arm and say "Look at this cut!". I cannot point to my wounded leg and say "Look at this kick!". I'd have to say "Look at the wound that the kick left!", or "Look at this bruise!", or whatever.

I know; I heard the crash.

But "crash" is an action here, not a result. Anyways, I'm not sure about "crash" though. One can look at a car crash and compel their friend to look at the "crash" (the remains/result of the crash(ing)), but not with a punch/kick/shake. "smash" and "crush" seem similar to "crash", but still strange, but perhaps just because people talk about "crashes" more often.

I think it leaves something to be desired.

Okay. I suppose that you would say that there's no division in the second group at all then?

I think referring to the nouns here as tools is too specific, and more likely a subset of a larger category. I would describe the nouns in this category as simply being items that are closely associated with the action, and/or items frequently used to perform the action.

Yeah, you're right. One can hammer something with a rock, or pin someone with their arms, after all.

Thunder could be either 1 or (3), i.e. a noun closely associated with the verb, or a direct result of the verb.

Well, "thunder" as a verb reminds my of other verbs of speech, like "talk", "shout", "growl", etc, but "thunder" as a noun seems to refer specifically to a certain phenomenon of weather, instead of anything that "thunders".

chain

Nah, "chain" was a noun first. Cf. Latin caténa

The verbs in (3) I think are just a subset of 2. For the most part, don't really leave any result that is clearly a product of the verb in question, so the "result" so to speak is the action at moment it occurs, which is the only time the action is clearly identifiable.

Yeah, that's what I mean by ephemeral. It's just inherent to the action that it's ephemeral, so a result is not easily perceived, so such verbs don't really follow a separate pattern from the second group. (Which is why I didn't label them as part of a third group.)

[linguoboy]

I know; I heard the crash.

But "crash" is an action here, not a result.

See, this is the distinction I'm not getting. The action is the object falling to the floor; the sound that reaches my ears when that happens is a result of the action.

And what about "punch" anyway? If someone has punched my punchcard five times, then I've got five what in it?

[Pthagnar]

holes

[Thry]

punchations

[linguoboy]

holes

A "hole" is not the same as a "punch". If someone shoots a bullet through my punchcard, it will have six "holes" in it, but still only five "punches".

[Thry]

punchhole?

[Salmoneus]

holes

A "hole" is not the same as a "punch". If someone shoots a bullet through my punchcard, it will have six "holes" in it, but still only five "punches".

It won't have any punches in it. [Unless it's capable of giving or receiving five more punches...]. It'll have five holes, or specifically punchholes. "Punch" can't be used for results - at least IMD. And the fact that there seems to be a dialectical difference in whether it can be used for results, but not whether it can be used for actions, demonstrates that there is a difference between actions and results.
Your 'crash' argument is ingenious, but doesn't work for verbs like 'kick'. A kick is always an action, never its result - whereas a cut is either (and more often the result, I think). So there is a difference there.

There's also a bit of a distinction in what we're calling an 'action'. Some verbs have their nouns refer to the action extensionally, others intensionally - "a ponder", for instance, is purely extensional (two ponders are distinguished by when and where and (by?/with?/in?/for?) whom they take place), whereas "a speech" is almost entirely intensional (two speeches are distinguished by their content, and you can even deliver the same speech in two different places and times, and two people can deliver the same speech, even). [Verbal nouns formed by -ing seem (mostly? all?) to be extensional, whereas -ion nouns seem to be intensional?]

It should be noted that derivation is never entirely predictable, and certainly not in English. For instance, -ion nouns are normally actions, or sometimes results. But "location" isn't the result of locating something - it had a location before you located it! It's... the thing discovered or learnt through the action. Whereas 'perdition' is theoretically an action or a result, but is almost always treated as a location.

And sometimes the categories aren't clearly distinguished. "Walk", for instance - not in the sense of going out for a walk, but in the same of a pathway along which one walks (and likewise 'promenade') - is walk the action (only an action that needn't be happening at the time), or a result (it's where people have walked), or a tool (by which one goes for a walk)?

At least in English, I don't think we can do much better than very vague tendencies for any sort of derivation (other than the most specific and rare).

[zompist]

I'd say that discussion is confused more than helped by talking about "results".

"Cut, break, fold, scratch, crack" all describe a deformation of the object, and moreover one that's characteristic of that action. A cut is instantly identifiable as being caused by cutting. Kicking doesn't produce such an identifiable deformation.

To add to the list: bend, mark, knot, twist, slit, wrinkle, dent, nick, pit, bruise, warp, turn, kink, gouge, stain.

I think linguoboy's example of the punchcard shows that this can be contextual— it's perfectly acceptable to say that the card has five punches in it, even if a circular hole or depression in general can't be called a punch. It wouldn't surprise me that nonce uses of other words could be devised. Perhaps a medical examiner can identify bruises caused by kicks, in which case he could say "These kicks here must have been produced by the left foot."

"Smash" certainly produces a deformation, but it encompasses the whole object, which perhaps is why we don't say that a car has a smash in it.

More mysterious is why we don't say "a pierce", which would seem to fit with the other words.

[Melteor]

We do have "pinprick" tho. Also "pricker".

"That's a cool origami."
"I origamied some shiny paper. Cool huh?"

[Radius Solis]

I'd say that discussion is confused more than helped by talking about "results".

This discussion, probably. But result nominalizations are a Thing, and it doesn't always involve identifiable deformations to an object. For example "song" derives from a PIE resultative form of "sing" (the other big example of this ablaut that we still have is knit -> knot).

There's several basic types of nominalization, three of which are action, result, and argument. All of these can be divided into subtypes - argument nominalizations can be broken down into agent nominalizations, patient nominalizations, instrument nominalizations, and so forth; our -er covers agents and instruments. Action nominalizations can be broken down into nouns giving an instantiation of the action ("a walk") and those that don't, i.e. participles and gerunds; some languages instead make aspectual distinctions in this category, such as imperfective versus perfective action nouns. And result nominalizations can be broken down into changes to existing things (as in the deformations) versus new things created (as with "song", maybe also "crash"), although I can't recall offhand if there are any languages that have different morphosyntax for result nominalization subtypes. We certainly don't.

'Feed' fits in category 1, because the noun is closely associated with and frequently used to perform the action in question.

It fits only imperfectly. The verb "feed" is less easily clear as to what grammatical roles its arguments have: the noun "feed" could be considered to signify the direct object of the verb rather than an instrument.

- I feed my chickens corn. / I feed corn to my chickens. <-- this pattern casts corn, i.e. the "feed", as the direct object, with the chickens being an indirect object - which you can't do with the other list-1 verbs. On the other hand:

- I feed my chickens with corn. <-- this cast the corn as an instrument, and you can use this pattern for any of the list-1 verbs.

[Terra]

See, this is the distinction I'm not getting. The action is the object falling to the floor; the sound that reaches my ears when that happens is a result of the action.

So the crash is the sound to you? ... That does sound good actually. Perhaps I was too fixated on the sense of crash in "car crash", that is, the debris left over.

And what about "punch" anyway? If someone has punched my punchcard five times, then I've got five what in it?

Punches. But if Superman punched (with his fist) through a wooden plank, can we say then that the plank has a punch in it? No, it has a hole in it. But, I think Zomp pointed out the right criterion: A punch from a fist does not leave a readily easily-identifiable result, but a punch from a hole-puncher does.

"Smash" certainly produces a deformation, but it encompasses the whole object, which perhaps is why we don't say that a car has a smash in it.

So, the problem is that the deformation is *too* complete? Okay, but that makes me wonder why would wouldn't refer to a crushed car as a "crush" instead of still a "car".

Going on a side meaning of "(have a) crush", meaning "infatuation", "crush" as a noun can mean either "the person that they have infatuation with", or simply "infatuation". (This may be an Americanism?)

What about "circle"? I can circle something with my pencil, but what I end up drawing is also a circle. It doesn't seem to adhere to either category. Oddly, "encircle" seems to mean something slightly different from "circle" as a verb; The former means to surround, while the latter means to go around in the shape of a circle.

[linguoboy]

More mysterious is why we don't say "a pierce", which would seem to fit with the other words.

Perhaps because of the prior existence of the concrete noun piercing? This generally blocks the derivation of synonymous nouns, except as nonce formations.

[Terra]

Perhaps because of the prior existence of the concrete noun piercing? This generally blocks the derivation of synonymous nouns, except as nonce formations.

But "piercing" is a recent coinage, isn't it? Why did it dominate over "pierce"? Deriving a noun result from a verb is at least 500 years old, if you look at when "cut" became a noun. (http://www.etymonline.com/index.php?ter ... in_frame=0)

[Radius Solis]

See, this is the distinction I'm not getting. The action is the object falling to the floor; the sound that reaches my ears when that happens is a result of the action.

So the crash is the sound to you? ... That does sound good actually. Perhaps I was too fixated on the sense of crash in "car crash", that is, the debris left over.

Eh? When we talk about a car crash, even when pointing at looking at the debris, we are referring to the whole event, not the debris remaining. One way you can tell: if you leave the debris there for a long time you then have to refer to the crash in the past tense: "wow, that must have been a really bad crash", not "wow, that is a really bad crash". And certainly not "*wow, look at all this crash still sitting around."

[linguoboy]

Perhaps because of the prior existence of the concrete noun piercing? This generally blocks the derivation of synonymous nouns, except as nonce formations.

But "piercing" is a recent coinage, isn't it? Why did it dominate over "pierce"? Deriving a noun result from a verb is at least 500 years old, if you look at when "cut" became a noun. (http://www.etymonline.com/index.php?ter ... in_frame=0)

According to the OED, piercing--as both an action noun (1390) and a noun of result (1400?)--is attested before pierce as a noun of either sort (1583, 1665).

[zompist]

"Smash" certainly produces a deformation, but it encompasses the whole object, which perhaps is why we don't say that a car has a smash in it.

So, the problem is that the deformation is *too* complete? Okay, but that makes me wonder why would wouldn't refer to a crushed car as a "crush" instead of still a "car".

Who knows? It just seems that it's easy to find words that refer to a defect, harder to find words that refer to the entire object as the deformation.

But there are some! "Wreck, ruin" work that way.

[clawgrip]

It's like you said, something that has been smashed does not have a deformation that is characteristic of smashing. It could just as easily have crumbled or exploded, or been crushed, carefully shattered, etc.

[Salmoneus]

So as well as action nominalisations, which seem to be universal, we have what I guess you'd call factitive nominalisations, where something comes into existence as a result of the action (crash (the sound), cut (the deformation or wound), etc), and patient nominalisations (feed, crush in the sense of be infatuated with). [Though I note that both of these examples have patients that aren't actually affected by the action - are there any semantically transitive examples out there?] And there are instruments (hammer, skate). And I think there are locations - most obvious is 'drive', also 'promenade', maybe 'jump' in the sense of a racing track.

Two interesting ones come to mind:
- draw. A draw isn't an act of drawing. Oh, wait, no, it is, just not for the primary meaning. OK.
- look. Primarily an action noun, but there's something else. How would you classify the expression on a person's face as they look at something? ["The look of love... he gave me a look... there was a terrible look on his face"]. This is probably also distinct from the... I don't know what you'd call it, the informative? As in 'it has a look of rain' or 'that dog has the look of a champion runner' - where 'look' means 'what it has that lets you know something when you look at it' (similar to the locate>location derivation in that sense).

[clawgrip]

Jump can be used for any object or location where people or animals are expected to jump. Buffalo jumps, for example, or, say, ramps in a video game are as likely as not to be called jumps.

[Ser]

Is not knowing what pattern to follow a common stumbling block for L2-learners?

Dunno, I think I just never assumed it was productive at all. Zero derivation is just one of many possibilities available anyway, see the use of the suffixes -ing, -ment, -ion/tion/ification/ization, -ance/ence, -er, -ate/ite, -ant/-ent, -age, -ery...

[Terra]

Jump can be used for any object or location where people or animals are expected to jump. Buffalo jumps, for example, or, say, ramps in a video game are as likely as not to be called jumps.

Indeed.

Two interesting ones come to mind:
- draw. A draw isn't an act of drawing. Oh, wait, no, it is, just not for the primary meaning. OK.
- look. Primarily an action noun, but there's something else. How would you classify the expression on a person's face as they look at something? ["The look of love... he gave me a look... there was a terrible look on his face"]. This is probably also distinct from the... I don't know what you'd call it, the informative? As in 'it has a look of rain' or 'that dog has the look of a champion runner' - where 'look' means 'what it has that lets you know something when you look at it' (similar to the locate>location derivation in that sense).

You mean "draw" as in playing a game? Yeah, that seems right. The main sense seems to be weird in the same way as "pierce". It nominalizes to "drawing" instead of "draw".

What about "loop"? It doesn't seem to be an instrument, but something looped isn't a "loop" either. Compare "lasso", which definitely is an instrument, and again, something lassoed isn't a lasso. Maybe "draw/drawing" is just weird in that it's main argument is the thing drawn, instead of the thing drawn *on*. This is all so very confusing...

Relevent: http://xkcd.com/804/

And I think there are locations - most obvious is 'drive', also 'promenade', maybe 'jump' in the sense of a racing track.

It's strange that one can "drive a car" and "drive a road", but only the latter can produce a patient nominalization.

"sign" too. When asked what one signed, one can respond either "the book" or "my autograph and a witty joke". If the book is a patient, then what's the latter?

It fits only imperfectly. The verb "feed" is less easily clear as to what grammatical roles its arguments have: the noun "feed" could be considered to signify the direct object of the verb rather than an instrument.

- I feed my chickens corn. / I feed corn to my chickens. <-- this pattern casts corn, i.e. the "feed", as the direct object, with the chickens being an indirect object - which you can't do with the other list-1 verbs. On the other hand:

- I feed my chickens with corn. <-- this cast the corn as an instrument, and you can use this pattern for any of the list-1 verbs.

Yeah, strange. What other verbs act this way?

[Herr Dunkel]

What's worse is the bisyllabic stress shift words: /prəˈdjuːs/ verb vs. /ˈprɒdjuːs/ noun.