Do

[ruby sparks]

I think I see. So, are you saying that we limit the grammatical role of agent to those items we attribute agency to in this or that sentence?

If so, surely that seems to be getting very close to a tautology, because obviously, if we weren't attributing agency, we wouldn't give the item that grammatical role.

And you're saying this had to be........discovered, by some ancient Indian Lingo boffin or something, like it wasn't obvious.

Kidding. I guess you're just noting that we make the agent/non-agent distinction and that this distinction is universally present in all of our languages.

How about this one, 'the pebbles were thrown onto the beach by the waves'.

Maybe it's just me, but the waves don't seem to be attributed with much agency in that one. Is there room for that, would you say, or would you say that sentence attributes agency?

What I'm getting at is, are there exceptions? Are there ('valid') cases where we don't, in fact, make the distinction? Can we be 'agent-neutral' when it comes to something being actively causal (or as you put it, in 'commands')? Iow, is making the distinction universal or general?

This may or may not involve us saying what we mean by 'agency' of course. If we mean a capacity to choose, respond, decide or intend.....then, I'm not seeing that regarding those waves, even if they're in motion (are active).

 

[Copernicus]

What I am saying is the following. All languages are structured. They chunk words together into phrases and phrases into clauses. Simple clauses combine to form complex clauses and sentences. If you study the syntactic structure of language, these building blocks exist in all human languages without exception.

The point here is that there is a role assigned to noun phrases that we can call "Agent". Another universal fact about human language is that no simple clause can ever have more than a single agent noun phrase (although a single noun phrase can encompass more than one agent). And no simple clause can ever have more than one "instrument" (i.e. "with" phrase in English) noun phrase. Again. this is true of all languages. Panini discovered this fact about Sanskrit, but it holds true for English and every other language that linguists have ever come across. That is why I say that we are "hardwired" to recognize a distinction between volitional (agentive) causation and mere physical causation. Agents in simple clause structure can only command a single instrument.

So here is what Panini discovered about Sanskrit. In that language, you can say:

Rama shot the elephant with a bow.
Rama shot the elephant with an arrow.

Both "bow" and "arrow" play the role of an instrument that the agent controls, but the underlying event has the bow causing the arrow to hit the elephant. Suppose that you want to mention both in the same sentence. In that case, you can only say:

Rama shot the elephant with an arrow from a bow.

You cannot say:

Rama shot the elephant with a bow ... arrow.

That is, there is no way to put "arrow" in the role of instrument when "bow" occupies that slot in the simple sentence structure. Panini created a rule in his grammar of Sanskrit that would make this prediction work, but he gave no explanation of why that is the case. From the perspective of modern linguistics, where we have thousands of recorded languages to work with, this prediction holds true in every language. So we are talking about a fundamental property of the way human beings think. My doctoral dissertation attempted to explain why.

Noam Chomsky once famously said that "language is a window into the mind".

[YOUTUBE]UnyJvCJrRys[/YOUTUBE]

 

[ruby sparks]

What I am saying is the following. All languages are structured. They chunk words together into phrases and phrases into clauses. Simple clauses combine to form complex clauses and sentences. If you study the syntactic structure of language, these building blocks exist in all human languages without exception.

The point here is that there is a role assigned to noun phrases that we can call "Agent". Another universal fact about human language is that no simple clause can ever have more than a single agent noun phrase (although a single noun phrase can encompass more than one agent). And no simple clause can ever have more than one "instrument" (i.e. "with" phrase in English) noun phrase. Again. this is true of all languages. Panini discovered this fact about Sanskrit, but it holds true for English and every other language that linguists have ever come across. That is why I say that we are "hardwired" to recognize a distinction between volitional (agentive) causation and mere physical causation. Agents in simple clause structure can only command a single instrument.

So here is what Panini discovered about Sanskrit. In that language, you can say:

Rama shot the elephant with a bow.
Rama shot the elephant with an arrow.

Both "bow" and "arrow" play the role of an instrument that the agent controls, but the underlying event has the bow causing the arrow to hit the elephant. Suppose that you want to mention both in the same sentence. In that case, you can only say:

Rama shot the elephant with an arrow from a bow.

You cannot say:

Rama shot the elephant with a bow ... arrow.

That is, there is no way to put "arrow" in the role of instrument when "bow" occupies that slot in the simple sentence structure. Panini created a rule in his grammar of Sanskrit that would make this prediction work, but he gave no explanation of why that is the case. From the perspective of modern linguistics, where we have thousands of recorded languages to work with, this prediction holds true in every language. So we are talking about a fundamental property of the way human beings think. My doctoral dissertation attempted to explain why.

Noam Chomsky once famously said that "language is a window into the mind".

[YOUTUBE]UnyJvCJrRys[/YOUTUBE]

Thanks. I always enjoy reading you talking about your subject of expertise. I always learn something.

So...you are not, I think, saying that every clause has to have a grammatical agent, but only that if it has, it obeys certain 'rules' (which reflect deeply on the way our minds work)?

In general terms, I have a huge respect for linguistics. I agree that it is a 'window to the mind'. Very much so. Not that I know in what ways, that's beyond me, but I still agree in principle. In this case, I'd be guessing that language is, perhaps amongst other things, revealing something about how we conceive of causality.

Could you comment on or deconstruct the sentence, 'the waves threw the pebbles onto the beach' in regards to whether you would say it has or implies an agent or not?

I note that I have mixed my metaphors by implying seeing through (window) and reflecting (mirror), but since I'm not sure which is best, I'm going to let it go for now.

 

[fast]

What I am saying is the following. All languages are structured. They chunk words together into phrases and phrases into clauses. Simple clauses combine to form complex clauses and sentences. If you study the syntactic structure of language, these building blocks exist in all human languages without exception.

The point here is that there is a role assigned to noun phrases that we can call "Agent". Another universal fact about human language is that no simple clause can ever have more than a single agent noun phrase (although a single noun phrase can encompass more than one agent). And no simple clause can ever have more than one "instrument" (i.e. "with" phrase in English) noun phrase. Again. this is true of all languages. Panini discovered this fact about Sanskrit, but it holds true for English and every other language that linguists have ever come across. That is why I say that we are "hardwired" to recognize a distinction between volitional (agentive) causation and mere physical causation. Agents in simple clause structure can only command a single instrument.

So here is what Panini discovered about Sanskrit. In that language, you can say:

Rama shot the elephant with a bow.
Rama shot the elephant with an arrow.

Both "bow" and "arrow" play the role of an instrument that the agent controls, but the underlying event has the bow causing the arrow to hit the elephant. Suppose that you want to mention both in the same sentence. In that case, you can only say:

Rama shot the elephant with an arrow from a bow.

You cannot say:

Rama shot the elephant with a bow ... arrow.

That is, there is no way to put "arrow" in the role of instrument when "bow" occupies that slot in the simple sentence structure. Panini created a rule in his grammar of Sanskrit that would make this prediction work, but he gave no explanation of why that is the case. From the perspective of modern linguistics, where we have thousands of recorded languages to work with, this prediction holds true in every language. So we are talking about a fundamental property of the way human beings think. My doctoral dissertation attempted to explain why.

Noam Chomsky once famously said that "language is a window into the mind".

[YOUTUBE]UnyJvCJrRys[/YOUTUBE]

Thanks. I always enjoy reading you talking about your subject of expertise. I always learn something.

So...you are not, I think, saying that every clause has to have a grammatical agent, but only that if it has, it obeys certain 'rules' (which reflect deeply on the way our minds work)?

In general terms, I have a huge respect for linguistics. I agree that it is a 'window to the mind'. Very much so. Not that I know in what ways, that's beyond me, but I still agree in principle. In this case, I'd be guessing that language is, perhaps amongst other things, revealing something about how we conceive of causality.

Could you comment on or deconstruct the sentence, 'the waves threw the pebbles onto the beach' in regards to whether you would say it has or implies an agent or not?

I note that I have mixed my metaphors by implying seeing through (window) and reflecting (mirror), but since I'm not sure which is best, I'm going to let it go for now.

"Threw" is a metaphor. No agent. No agency.

 

[ruby sparks]

Not sure. In grammar terms, it's a verb. To throw is to propel something. A metaphor is a noun?

 

[ruby sparks]

Or, perhaps, look at this:

1. The dog dug up an old bone.
2. We dug up some interesting facts.

The verb in the second one. Is that a metaphor verb? Maybe I was wrong to say a metaphor has to be a noun.

Either way, the second one would involve a metaphor and an agent. So I'm not sure a metaphor means no agency.

 

[Copernicus]

Not sure. In grammar terms, it's a verb. To throw is to propel something. A metaphor is a noun?

Verbs and nouns are grammatical categories. Metaphors are not grammatical categories. They are essentially a type of analogy--a technique for describing one concept in terms of its similarity to a different (usually more familiar) concept. The word "throw" is a verb in the first sentence, and a noun in the second:

1. The pitcher threw a ball to the catcher.
2. The pitcher's throw fell short.

Usually, the verb "throw" entails movement of the arm, but there is a secondary sense where it just means "cause to move something through the air". So "The wave threw pebbles on the beach" can have a perfectly conventional literal meaning, but its conventional usage may derive historically from a metaphorical sense of "use the hand with arm motion to cause something to move through the air" where the "wave" represents a kind of metaphorical "arm". Quite often, literal usage evolves historically from metaphorical (non-literal) usage. When people turn a "live" metaphor into a repeated convention, it can "fossilize" into a grammatical structure. There are countless examples of this in the linguistic literature on metaphor.

Note that forces of nature are not normally conceived of as volitional agents. So you would not say "The ocean threw pebbles on the beach with a wave", unless you are thinking of the "ocean" as a kind of volitional entity.

Or, perhaps, look at this:

1. The dog dug up an old bone.
2. We dug up some interesting facts.

The verb in the second one. Is that a metaphor verb? Maybe I was wrong to say a metaphor has to be a noun.

Either way, the second one would involve a metaphor and an agent. So I'm not sure a metaphor means no agency.

You really ought to read Lakoff and Johnson's  Metaphors We Live By. It is all about the process of how metaphors become idioms that "die" or gradually lose their semantic transparency and become fossilized linguistic structures.

 

[ruby sparks]

You really ought to read Lakoff and Johnson's  Metaphors We Live By. It is all about the process of how metaphors become idioms that "die" or gradually lose their semantic transparency and become fossilized linguistic structures.

That was once recommended to me by a philosophy professor with a special interest in metaphors who was running a Sunday morning philosophy Club that I attended. I meant to buy it, but never got around to it, though I read up on Lakoff and read some free online material by him, and was fascinated.

 

[fast]

Or, perhaps, look at this:

1. The dog dug up an old bone.
2. We dug up some interesting facts.

The verb in the second one. Is that a metaphor verb? Maybe I was wrong to say a metaphor has to be a noun.

Either way, the second one would involve a metaphor and an agent. So I'm not sure a metaphor means no agency.

That's what I get for talking in incomplete sentences, lol.

Where there is no agent, there is no agency.

The ocean is not an agent. The ocean has no agency.

If someone throws something, there is an agent and thus agency.

If something is thrown and there is no agency, then "throw" becomes ambiguous.

If the alternative usage becomes mainstream such as you do not imagine the legs of people when discussing the legs of a chair or table, then even though one is a metaphor for the other, it becomes a lost tale of etymology, as the alternative use is collectively used by fluent speakers without invoking its history.

 

[Copernicus]

Or, perhaps, look at this:

1. The dog dug up an old bone.
2. We dug up some interesting facts.

The verb in the second one. Is that a metaphor verb? Maybe I was wrong to say a metaphor has to be a noun.

Either way, the second one would involve a metaphor and an agent. So I'm not sure a metaphor means no agency.

That's what I get for talking in incomplete sentences, lol.

Where there is no agent, there is no agency.

The ocean is not an agent. The ocean has no agency.

If someone throws something, there is an agent and thus agency.

Agency is a very complicated concept, so it is easy to get confused by its different uses in discussions like this. What I have been talking about above has to do with the way a human mind organizes reality into linguistic units such as simple clause structure. A hard and fast rule across all languages is that simple clauses contain noun phrases that play semantic-syntactic "roles". There can only be one grammatical Agent in a simple clause and one grammatical Instrument (immediate physical cause). When we talk about causal chains, we organize the description around causal verbs that can have only one Agent when an agent is expressed. And, if the agent is expressed, it must be the subject of the sentence (in active voice). There are no languages that violate this generalization. If the Agent isn't expressed, then the subject of the causal verb is the physical causer (Instrument), if it is expressed. We are talking about universals of language that define how the mind translates thoughts into language. So these are important observations for any Theory of Mind that philosophers come up with. Unfortunately, most philosophers, not being trained in linguistics, are not really aware of these universals of clause structure, although a few are. (Dennett, for example, is not.)

The human mind has evolved to build elaborate causal models of reality. Those models let us make reasonable predictions about future events, and those predictions ensure and prolong our survival. Hence, agency and instrumentality are important aspects of reality to us. Language structure has evolved very sophisticated mechanisms for describing causal chains. Some verbs encapsulate extremely complex causality--e.g. "shoot", "throw", "collide", "eat", "defrost", etc. If you analyze the causal interactions that those verbs express, then you discover that grammar limits how we can describe the underlying concepts. Those verbs can always take volitional agents as subjects, but they can also take non-volitional agents. The catch is that volitional agents always take precedence over non-volitional ones. So you can say "John shot the deer with this gun" or "This gun shot the deer", but, if you happen to mention both the Agent "John" and the Instrument "gun" in the same simple clause, then "John" must be the subject: "John shot the deer with this gun." The Instrument "gun" cannot be expressed as the subject when an Agent is present to knock it out of its potential to occupy the grammatical subject slot.

If something is thrown and there is no agency, then "throw" becomes ambiguous.

But, is "catapult" the agent in "The catapult threw rocks into the fortress"? It is a non-volitional agent of sorts, but it is not an Agent in a linguistic sense. As a non-volitional force in the complex causal chain, it plays the role of Instrument in the sentence. The human mind makes a fundamental distinction between causation under volitional control and causation not under such control. For a true Agent subject, you need a sentence like "The attackers threw rocks into the fortress with a catapult".

If the alternative usage becomes mainstream such as you do not imagine the legs of people when discussing the legs of a chair or table, then even though one is a metaphor for the other, it becomes a lost tale of etymology, as the alternative use is collectively used by fluent speakers without invoking its history.

Exactly right. Metaphors can become linguistic conventions that can become detached from their conceptual origins. We've all seen folks getting into heated discussions over what words mean. To the extent that posters in internet forums forget your point, they tend to engage in etymological fallacies.

 

[Speakpigeon]

On the other hand, I think they over-simplify. Whilst we probably do not have free will in that sense, we do have certain sophisticated capacities for agency (using the term 'we' generally, because the capacities will vary from person to person, as well as between species, as well as between biological organisms and non-biological machines).

As such, I'm in favour of debunking popular conceptions of free will, but not of throwing the baby out with the bathwater.

Most recently I came across this article:
https://www.sheffield.ac.uk/polopoly_fs/1.101516!/file/smilansky-on-shallowness.pdf

Yes, Smilansky seems close enough to my own views...
EB