These are some notes which I wrote down in the last few months. I was thinking of writing a book, so I wrote in order to make the topic of my book a bit more clear. This is the only use of this text – to make it clear for myself what I am supposed to write about in my book. The only result of writing everything found here, which I would acknowledge, is that the title of my book should be ‘Reason, Language and Actions’. Nothing else should be taken seriously, nothing is worth mentioning etc. I did not proofread the text for grammar mistakes. If you want to take a look at it, you’re on your own.
Let me shortly retrace the cognitive steps which which led me to writing this. It all started with the later Wittgenstein, whose writings enforced my conviction that the entry door to any other philosophical problem was philosophy of language, partly due to its propedeutical role (one must make sure one is asking a meaningful question before looking for an answer) and partly due to its irreplaceable role as a tool for conceptual analysis (my inherited view being that philosophy must deal with the analysis and clarification of our concepts).
However, several Wittgensteinean arguments, among which the rule-following considerations seemed the most important to me, supported the idea that a (normative) theory of language is not possible. A related point which I got from Wittgenstein was that any attempt to map the relation between language and world as some sort of isomorphism was irredeemably flawed.
Thus I became convinced that the connection between our language and the world around us must be accounted for in a different manner – not as some sort of correspondence between words and sentences floating in logical space and “metaphysically real” objects and states of affairs, but as an intertwinement of our utterances with other (non-communicative) actions within our shared practices, which we perform while being a part of our (natural) environment.
The main problem with this approach was that logic (and reason) had no place in it. There was a clash between “the space of nature” and “the space of reasons”. I was familiar with other attempts to solve this problem, but they did not seem satisfactory. I have pursued my own attempt to bridge the gap between communicative and non-communicative actions and place them back in some sort of logical space, but the more I was trying to overcome various conceptual difficulties, the more I was aware of the deep nature of the original problem. It was always as if there where two vocabularies – the naturalist one and a second vocabulary, used to talk of persons, actions, intentions, responsibility and norms (including logical rules).
So, how does an action fit in a logical space? To make it fit, we first have to conceive the logical space in a different way – namely, as the space in which one traces responsibilities (or commitments). Uttering A makes you responsible for having fulfilled the necessary conditions for the successful utterance of A. Thus, by making the assertion that Socrates is a man, for instance, you commit yourself to the assertion that there exists at least a man (since Socrates is a man only if there exists at least a man). You also commit yourself to the assertion that you have information that Socrates is a man (since one can sincerely assert something only if one has information about it).
Now, what is a necessary condition for successfully asserting that A?
You can assert A without asserting all the logical consequences of A, but if you assert A and reject one of its logical consequences you are performing some opposing actions. You can be perhaps successful in performing both assertions (or at least the first one), or you can be taken as such, but if both were part of the same series of actions one will fail to see what are you trying to achieve.
(You can assert something which you do not believe, but… – is this a similar case?)
You can also make an uninformed assertion. This is going to be weird. You had a responsibility to get information before saying something (“The king of Tuvalu is blind”). If you did not fulfill that responsibility, one could claim that you did not make a proper assertion. You said something, but you asserted nothing (by contrast, you can reject a proposal without getting any (additional) information – “no, I cannot come to the movies tonight”).
If I say that the king of Tuvalu is blind, I can claim that I’ve made an assertion, but my claim should be rejected.
In the same way, if I claim to having made a tea by putting some tea leafs in a cup of cold water, my claim should be rejected. This is not how one makes a tea.
It could be said: “This is not how one makes an assertion”. It seems that the concept of “assertion” (like the concept of “making tea”) is involved here in a way in which it is not involved in the “Socrates is a man” case. Effectively making the assertion that there exists at least a man (or even believing it) is not required for the success of my assertion. However, my commitment to the existence of at least a man seems to be required. If I say “Socrates is a man, but I do not believe that there exists at least a man”, it should be rejected that I have made the assertion that Socrates is a man.
Thus, sometimes the necessary conditions of my actions are other actions (gathering information, boiling the water etc.), while other times they are commitments (but not the actual performance of some commissive speech acts).
A further complication. What if I say that Socrates is a man but he does not exist? It might be said that I fail to make the proper assertion that Socrates is a man, although in a different way. That Socrates exists is a necessary condition for him being a (real) man. I assume, for now, that a proper assertion cannot be made in a fictional discourse (so one cannot actually assert that Sherlock Holmes did not exist). So the necessary condition is a condition for the truth of what I am saying, but not directly for my action of asserting that Socrates is a man. It is a condition for my assertion only due to the rule that one cannot assert something false (if something is impossible, one cannot promise to do it; and one cannot do it as well).
So the content of my action is such that I cannot perform the respective action. If a door is closed, one cannot close it. “You cannot close a closed door” is similar to “you cannot toss a non-existent ball” and to “I cannot assert something about Sherlock Holmes” (which I can only say in fictional discourse as well).
“You cannot make tea using dirt” – this is also due to the nature of the concept of “making tea”. Still, it seems that “You cannot close a closed door” displays a more general rule: “If the result of an action A is state S, you can perform A only if S is not the case (we ignore time specifications at this point).
(sometimes we call actions by their results, other times by a description of what is performed: shooting vs. pulling the trigger; unplugging x vs. pulling x’s plug; sometimes it is hard to distinguish between the two: opening the door)
Let us compare:
(1) You can unlock a door only if you get a key (its key). / You can make tea only if you boil water.
(2) You can unlock a door only if it is locked.
What am I trying to do here? I am trying to figure how can we talk about logical relations between actions. The initial proposal was to say that B is a logical consequence of A IFF B is a necessary condition for the successful performance of A.
The first problem was that uttering (performing the assertion) that some men are mortal did not seem to be a necessary condition for uttering (performing the assertion) that all men are mortal, in the way in which gathering reliable information that all men are mortal seemed to be a necessary condition for asserting it.
“I cannot sit in a chair (intentionally sit myself in a chair) if I do not believe that there is a chair” – “I cannot assert that all men are mortal if I do not believe that all men are mortal” – are these two on a pair?
“But what is it to believe that P?” – this could make me loose track again.
I cannot assert that A without committing myself to the truth of A. And I cannot commit myself to the truth of A if A is false (in the same way in which I cannot fly a wet kite).
“Is committing oneself to the truth of P some sort of action?” – If so (and we are not psychologists), it results that by uttering P to make the assertion that P one also makes an infinite number of other actions (committing oneself to the truth of P & P, to the truth of P & (P & P) to the truth of P only if P, Q only if P a.s.o.
Perhaps I should distinguish between the logical space conceived by the mathematician and the “space of reason”. It does not seem a reasonable requirement to commit to the truth of “P only if P, only if P” when you assert that P.
It does seem reasonable, however, to commit to the truth of “Some men are mortal” when you assert that all men are mortal. Let me ask myself, now: am I supposed to describe what I have done by saying that all men are mortal (in the appropriate context) by saying that I have asserted that all men are mortal, but also that some are and that no immortal is a man as well?
I do not think so. Let me try again:
(a) I assert that no immortal is a man.
(b) I assert that all men are mortal.
My first intuition is that one can do (b) without doing (a). This is not only because the two sentences – (a) and (b) – have different conceptual content (so to say). “Pulling the trigger” and “shooting” do not have the same meaning, but except for some special cases (the gun was not loaded, I have pulled the trigger but the gun did not shoot etc.), in doing one of the two one is also doing the other one. Here, when one does (a), one does not usually do (b).
Another analogy: putting an object in the cupboard vs. removing it from the desk (but putting men in the class of mortal beings I remove the from the class of immortals, such that no immortal can be a man; we must of course suppose that an object can only be placed on the desk or in the cupboard).
So perhaps my second intuition could be: by doing (b) I also do (a).
Still, what about (c) denying that any immortal is a man? Doing (a) is “logically equivalent” with doing (c) in an even more straightforward way, but asserting and denying are different speech acts. How can one perform both at the same time?
“I have denied that any immortal could be a man. – No, you did not do that. You just asserted that no immortal is a man.”
Could someone talk like this? Why not? Perhaps what the second person has in view in this dialogue is: “The act of denying P must be opposed to the act of asserting P. You cannot deny that P if it wasn’t asserted that P. Nobody said that an immortal is a man, so you could not have denied that.”.
To this the first person could reply: “An act can be opposed to another act not only by cancelling it or reversing its result, but also by preventing it. By asserting that no immortal could be a man I have prevented the assertion that an immortal is a man. So my assertion was also opposed to that assertion, so I am right to call it also a denial.”
What is the clash about? It has to be about two views: according to the first, actions are concrete particulars; they occur in time and exclude each other. If A happens at t, no other action can happen at t and come down to the same event as A. According to the second, talk about actions is talk about abstract objects. Actions are similar to sentences. They float in a logical space. Performing an action is something which reverberates through the whole logical space.
So I am back to square one. Or maybe not. Let us say that (a) and (b) are different particular actions and see where this leads us. The content of (a) follows logically from the content of (b). If an object can only be placed either on the table or in my hand, by removing it from the table I place it in my hand. Now, suppose I say that what I did was to remove the object from the table. I could not have done that without placing it in my hand. But I insist that I did not do that second action. (“I could have asserted that all man are mortal without asserting that no immortal is a man.”)
Perhaps I say: “do not look at my hand, look at the table, I have removed the cup from it, see?”
Then, the person preparing a tea could also say: I did not boil the water, I have made tea. Do not look at the electric boiler, look at the cup of tea!
In the second case, boiling the water is only a part of the action of making tea. So I reject it as an action (performed by me) on the ground of its being only a part of the action I have performed. In the first case, however, I do not want to say that placing the cup in my hand was a part of the action of removing the cup from the table so that is why it was not a distinct action performed by myself. It was not a part of anything. It is not something I have done. I did the other (logically equivalent) thing.
What if we do not have a logically equivalence relation? Could it not be said that asserting that some men are mortal is a part of asserting that all men are mortal. Suppose I say that all Romanian presidents were men. Isn’t the assertion that Ceausescu was a man part of this?
“Ok, but boiling the water was a part of the physical event of making the tea, while, asserting that Ceausescu was a man was not a part of the physical event of asserting that all Romanian presidents were men. (in the way in which saying the words ‘Romanian presidents’ is a part of saying the sentence ‘All Romanian presidents were men’.”.
So this is why one might be reluctant to accept that the logical consequences of an utterance are similar to the “logical consequences of an action”. The logical consequences of P were not uttered by uttering P means “They were not a part of the physical event to which the utterance of P is reducible”.
“One can make wine only if one crushes grapes.” – Does this mean that Jesus did not make wine at the wedding in Cana, Galileea, after all? Suppose someone says: “One cannot miraculously unlock the door (without a key) for conceptual reasons.”. What should I say about that?
Jesus miraculously made wine out of water.
Jesus did not crush grapes.
One can make wine only if one crushes grapes.
Socrates miraculously asserted that all men are mortal.
Socrates did not assert that some men are mortal.
One can assert that all men are mortal only if one asserts that some men are mortal.
Are these cases similar? If they are, as I suspect, that belief in logical relations is similar to belief in miracles. That is, supposing that a proper assertion could not be (physically) made without first asserting all the grounds for your assertion.
Suppose I pour water in a cup, put tea leaves in it and offer you a tea. “What about boiling the water?”, you say. I reply: “Well, I will do that later. Now drink the tea while it’s hot.”. It sounds curious, but it should not be that weird. I can assert P now and assert P’s grounds later. This is the same. If by making tea I assume responsibility for boiling the water, it does not matter when I do it. Responsibility transcends time. I am bound to boil water, but it does not matter when I do it. In the same way, asserting that P binds me to the responsibility of asserting P’s grounds (or perhaps its logical consequences; and perhaps not necessarily only asserting them, but also agreeing to them when somebody else asserts them), but it does not matter when I do it.
“So by making wine Jesus was assuming the responsibility to crush some grapes?” – Indeed, if this is what means to make wine and this is what Jesus did. If ‘unlocking a door’ means ‘procuring the key of a door, putting it in the lock and turning it, such that the door can be opened etc.”, then even if I unlock the door miraculously, I am responsible for procuring its key, putting it in the lock and turning it. It’s either that, or it must be said that I did not unlock the door, but only “make it so that it can be opened”.
Suppose I want to perform a particular move in a game of chess. The move, let’s say, by which I turn a pawn into a queen. I could, of course, arrange the pieces such that I immediately make that move. Nothings stops me. It is not physically impossible for me to make this move. However, if I want to perform that move “in a game of chess” I must make other moves in order to get to that position. Suppose the rules of the game would allow me to perform the move at the beginning of the game. I would make the move immediately, but then I will arrange the pieces for the beginning of the game and try to make it so that I will reach that position. At that point, I would skip the move which I have already made at the beginning of the game. This might be awkward, but it would be perfectly in order for this version of chess. What is important is to note that by making the move at the beginning of the game I assume the responsibility for bringing it about that my pawn reaches in that position, such that I can make the move according to the rules of traditional chess.
Assertions are made according to similar rules (let’s say that for now). Most non-communicative actions are not. This is precisely because they are embedded in our natural environment. If there was no nature, no causation, but only responsibility, we could all perform miracles all the time. But we would still be responsible for doing a lot of things.
There are non-communicative actions for which the order of performance does not matter of course. It does not matter, for instance, in which order do I put on different clothes when I dress myself. But I still cannot fully dress myself first and put my shirt on only afterwards.
Also, if B could be done (as a part of doing A) after doing A, then B would not be a necessary condition for doing A.
Now, one could say: “The existence of at least a man is a necessary condition for Socrates being a man. It does not follow from this that uttering that at least a man does exist is a necessary condition for uttering that Socrates is a man. The first relation concerns facts (or sentences, if you wish). The second concerns events. You are making a terrible confusion here”.
To this I would reply that there are no pure facts. The existence of at least a man depends on what we say. If we did not call anything “man”, there would be no men. In that case, it would be impossible to assert that Socrates is a man, of course, but because it was never asserted that a single man does exist, since nobody was called a man. So asserting that a man does exist is a necessary condition for asserting that Socrates is a man, after all.
Suppose I invent a new word and I call myself a xorp. I am bound to assert (or accept) that at least a xorp exists. I am also bound to assert (or accept) all sentences specifying the content of the “xorp” concept. Moreover, if someone accepts that by saying “I am a xorp” I have made a proper assertion, I am bound to say what a xorp is to that particular person. And if I say, for instance, that a xorp is a bearded man wearing headphones, I am responsible (towards the same person) for asserting or accepting that I wear headphones. And if I take my headphones off, I am responsible for ceasing to call myself a xorp. So I would say that the following is an example of reasoning:
(1) Asserting that you are a xorp.
(2) Asserting that a xorp is a bearded man wearing headphones.
(3) Removing your headphones.
(4) Refraining from calling yourself a xorp.
What happens is that the performance of (1), (2) and (3) make me responsible for doing (4) [I take it that refraining from doing something is also an action]
In the same way, asserting (or agreeing) that there is at least a chair in room 1 follows logically from bringing a chair into room 1.
“According to this view, paying you 5 euros follows logically from promising to pay you 5 euros. Am I not responsible for paying you the money if I have promised to do so?”
Well, perhaps you are. The promise, however, is not a sufficient condition for the pay. It is not enough to promise something in order to do it. Promising that I will pay you a billion euros does not make me able to do so. Along the same lines, it could be said that the pay is not a necessary condition for the promise. A promise I did not keep was still a promise. In this case perhaps you talk about responsibility in a different way. My concept of responsibility was not one of moral responsibility. I could use it to say that one is responsible for all the actions one makes, but one is not morally responsible for all the actions one makes. For instance, no moral responsibility follows from drinking a cup of tea I have prepared myself in the privacy of my home. In my sense, however, I am still responsible for performing that particular action. This, in turn, means that I am responsible to asserting or agreeing that I am drinking something, to refraining from saying that I never drunk tea, to sipping tea from the cup and so on. This network of responsibilities, I would say, is what constitutes the content of the concept of drinking tea. In other words, it is because I enter into this network of responsibilities that what am I doing is drinking tea and not something else. Still, there is no moral responsibility in that.
“Aren’t you responsible to agree that you are drinking something because you have a moral responsibility not to lie to others?”
It is not that. If I were to describe what I was doing, I would be responsible to assert that I was drinking tea. This, however wouldn’t be due to any moral responsibility towards anyone, but due to the responsibilities one has when one undertakes the task of describing something. This, in it’s turn is only related to what it means “to describe something”.
“But isn’t this circular? I seem to say that the content of the concept of ‘drinking tea’, which is also the implicit conceptual content of your action of drinking tea depends on the concept of ‘giving a description of what one is doing’.”
I do not see any circularity here. The contents of all our concepts depends on the responsibility network into which our actions figure. That network itself depends on the way in which we try to adapt to our environment (by having social interactions) and on the laws of nature.
For instance, producing always a different sound while performing the same kind of action does not foster any useful social interaction, so this (roughly put and only in part) is why I am responsible for saying that I drink tea when I do it.
“Right, but when you say ‘this is why’, what do you have in mind, a cause or a reason?”
It’s the cause of my behaviour and the reason of my responsibility to act as I do.
“And now you’ve hit solid rock, since nothing can cause responsibilities and natural phenomena can not be reasons for anything.”
It is not the biological relation of parenthood which “causes” the responsibilities of a parent. To this one could reply that society pushes biological parents into the social role of a parent and the responsibilities come with that. But being “pushed by society” to take on some social role does not amount to undertaking a certain responsibility. One could act as a “parent because this is what society expects from me”, but still elude responsibility. Suppose this person says: “I act as a parent because this is what society wants me to do and it would be troublesome to act otherwise, but I do not think that I have a duty to act like I do”. Responsibility, it seems, cannot be caused, but only assumed.
It seems that we must see the assumption of a responsibility as some sort of act one can perform freely. Otherwise, that would be no responsibility, but only constraint.
“But isn’t the fact that I am responsible for raising my children the effect and my act of assuming the responsibility for raising my children the cause of that effect?”
What if I declare war to another country as ruler of my own country? One could, of course, say that the state of war between our countries is the effect of my declaration. The state of war, however, is not a “state of affairs”, so to speak. It could still be that nothing has changed. Your country must prepare to defend itself, so to speak, against any attacks coming from my country, but your government could say: “We do not think they would attack us. We are at war, of course, but we are perfectly safe as we are, so we do not have to change anything”.
“But they have to talk about it.”
Indeed. They could completely ignore my declaration, nevertheless. It would be disputable then that your country is at war with my country, but from my point of view we are at war.
As a side note, all declarations (as speech acts) seem to work like this. I cannot declare that a lump of gold is to appear on my desk, but I can declare that my friendship with you is over.
Not any social / personal state can be achieved through declarations. We cannot stop being married by declaring it.
Anyway, I can start being responsible for something by declaring it. (I am zero-level responsible for an action by performing it. So there is no need to declare anything like “I am responsible for the declaration that I am responsible for raising my children”).
Duty (not necessarily moral)-responsibility: comes down to your obligation to perform several actions / to get involved in certain activities (this could be a complex matter), or to refrain from doing certain things, to observe rules, to provide justifications and make amends when you fail to do any of the above etc.
Moral responsibility is a duty such that you acknowledge that failing to keep it would be morally wrong (besides the “wrongness” of evading a responsibility in itself).
I am not morally responsible for avoiding to harm another being for fun just because these kinds of actions are wrong, but because I assume this responsibility and acknowledge the it would be wrong to decline it (or fail to keep it).
One could call this personal morality.
From doing A does not follow that the agent of A had promised to do A. Indeed, since promising to do A is not a necessary condition for doing A (although it could be in some special cases – I cannot return a borrowed thing without having borrowed it and I cannot borrow a thing without promising to return it).
Also, from promising to do A does not follow that the utterer of the promise does A (although promising to do A tends to make people do A). The promise and the action might have the same conceptual content, but no logical relation.
This could be contrasted to doing A and asserting that you have done A. Under a strong concept of assertibility the performance of A is a necessary condition for the assertion that you did A. Here we have a logical relation (also, similar relations hold between the successful performance of an action and assertions that the regular effects of that action did occur (e. g.,opening the door and asserting that the door is open).
A command to do A is a necessary condition for the addressee of the command having performed the order that A, but not for the performance of A in itself (except for special cases – executing a prisoner cannot be done without an order etc.).
Not even the case that A can be performed is a necessary condition for uttering the order that someone does A (you cannot order that someone “should do” A or “must do” A, but only that someone does A).
“Why do you say that every action has an addressee? We distinguish between ‘John has opened the door’ and ‘John has opened the door for/to himself’.”
We do, and perhaps ‘John has done something for himself’ does not follow from ‘John did something (A)’ and ‘John has benefited from A’. One could benefit from an action done for somebody else. However, if the action has a purpose, the purpose should have a conceptual content (it should fall under some description). The agent could provide the conceptual content, of course, but we can distinguish between a basic description of an action and a description of an action which includes its purpose: pushing the button vs. turning on the lights (I will not enter into any details about this distinction right now. The agent can provide a basic description of the action she performs, but there is no point to describe the action’s purpose if there is no addressee.
“So the addressee is a necessary condition for the action’s purpose having a conceptual content?”
You can put it that way, if you want.
“Not every action has a purpose, though.”
Gratuitous acts, hmm,… I suppose that is what you have in view.
“Think about a person thinking at something. You’ll say that thinking is talking or writing, while what that person actually does is to simulate thinking, but, leaving aside such verbal differences, there are acts of thinking which do not resemble talking to somebody (not even to yourself). You might want to call such acts ‘no-intonation (since your not imitating speech) inner talking (or writing)’. Whatever the name, such acts (which can be caused by sensory input, of course) might not have any particular purpose.”
Well, you do not need to resort to inner acts in order to prove your point. One could jot a few random notes down impulsively, without aiming to achieve any particular purpose. However, what goes for intentions goes for purposes as well. We can talk of the intention with which an action was performed without any ontological commitment to some psychological object – the intention – which accompanied the action’s performance. An intention can be attributed at any time, perhaps in a reply to the question ‘Why did you do A?’. It might even be that attributions of intentions and purposes coincide. In any case, attribution of purposes can work in the same way as attribution of intentions. My action has a purpose if I can attribute a purpose to it at any point in time.
“But you said it yourself, the agent does not need to attribute a purpose to her actions.”
What I said was that the agent does not need to specify the conceptual content of her actions’ purposes. Nevertheless, I have no problem with a case in which I behave in a certain way (say, I jot down some random notes, which I might even delete afterwards – though deletion in itself does mean lack of purpose, it could have been a failed attempt to achieve a purpose), so I behave in a certain way and do not attribute any purpose to my behaviour. In such a case, I would say, what happened would not be that I have performed an action.
“So writing a random note which I forget about afterwards is just ‘human behaviour’ and should not be called an action? This seems forced.”
Ok, think about it this way. Suppose I do not want to say that I have no responsibility for what happened. I want to say that writing down those notes was my doing. Perhaps I wrote them on a notebook which was not mine etc. Would there be any point in talking about responsibility (even zero-level level responsibility, which is apparent in questions like ‘how did A?’, where A replaces a basic description of the action in cause) if nobody was affected?
“Well, as long as somebody could have been affected, speaking in principle, why not?”
Ok, I see the point. But then, even no-intonation inner thinking could affect someone, since it is in principle possible (i. e. conceivable) that somebody else would ‘telepathically hear my thoughts’ and so on. Somebody could be affected by ‘telepathically hearing my feelings’ as well, or by directly hearing my snoring, for that matter. That does not make me zero-level responsible for snoring. It was something which happened to me, not my doing. Perhaps it could have been prevented by me. If it was, I am responsible for not having prevented it, but not for snoring per se. Better examples could be easily constructed.
I do not assume zero-level responsibility for every movement of my body, even if that movement is part of an action performed by myself. Basic actions are those actions for which I can assume zero-level responsibility, but that is not determined by their playing a role in a causal chain leading to other persons, but by their playing a role in a series of actions leading to a purpose. Leaving aside the case when the series has only one element, each individual step in the series can be described by the role it has in achieving the purpose of the entire series, which makes each step describable in terms of a particular purpose it serves.
So if no purpose is in view, why call something an action?
“You make my task quite difficult. I cannot give you an example of an action lacking a purpose, since by being used as an example the actions in case would automatically have a purpose. Your use of the word ‘purpose’ is, however, quite lax. It would seem, according to a stricter view, that we only assume responsibility for what could be part of a story of our lives told in terms of means and purposes. But we include more in that story, provided that it remains narratively coherent. Why don’t you agree that narrative coherence is something larger and more integrative than the means-purpose fitting scenario?”
I do not know. Perhaps I have to think of Elizabeth Anscombe’s arguments for the idea that the means-purposes scenario is applicable to all our actions. Perhaps she sees the means-purposes relations as reducible to causal relations, while I do not see them like this.
“But why distinguish between ‘basic description of an action’ and ‘purpose description’ if every action has a purpose, then?”
Because they are different descriptions. It’s that simple. We can describe a process at an observational level and at a physical level, so to speak. All processes described at the first level will have a description at the second level, but that does not mean that we should not distinguish between the two kinds of descriptions. This might lead to the question ‘what is that which is described in those two ways’, but it need not do so. And if it does, this only means that my analogy stops right there.
“So, basicaly, what you’re saying is this. Any actions must be conceived as having an addressee, since the addressee is a necesary condition for the action’s purpose having a conceptual content. In addition, having a purpose (with a conceptual content) is a necessary condition for something being assumed zero-level responsibility for, which, in its turn, is a necessary condition for something being an action.”
Let me try to say it again. No action without an agent assuming zero-level responsibility for it. No assumption of zero-level responsibility for something which is not (at least) part of a series of actions and no action series without a purpose (thus making any part of the series have a purpose). No purpose without conceptual content. No conceptual content for the purpose of an action without an addressee of that action. Therefore, no action without an addressee.
“I still think that all the links in your chain of thought depend on an idiosyncratic understanding of words like ‘purpose’, ‘responsibility’, ‘action series’ and even ‘action’.”
I wouldn’t call it an idiosyncratic understanding. This is, after all, a conceptual proposal. It is not like we are trying to describe ‘action-facts’ and you are not satisfied with my description. You can be dissatisfied with my conceptual proposal if it seems to you that it complicates the way we think about actions or produces unnecessary problems (or philosophical puzzles) or if you favor another proposal more, for whatever reasons.
“And what are the benefits of your proposal?”
For one think, it encourages us to expand the “Whom am I talking to?” question into the “Whom am I doing this for?” question, thus uniformly treating all our actions, not only the communicative ones (also, by this step we might attempt to expand more concepts from the theories of communicative actions – i.e. speech acts – to concepts applicable to all human actions. It also makes it easier for us to get from ‘zero-level responsibility’ to ‘responsibility towards someone’ and distinguish ‘responsibility towards another person’ from ‘responsibility in front of another person’, since only in the first case must the other person be the addressee of our action.
One more time. The more general model for the relation of rational entailment starts with responsibility entailment. It goes like this. By doing A I am responsible for having done B.
“What if you only put the teabag in the hot water? This does not make you responsible for having boiled the water, of course.”
Sure, but I am not responsible for having made the tea by myself, in this case. Were I responsible for making the tea, I would have been responsible for having boiled the water. If I co-author a book, I am not responsible for having written every single sentece in it. I might take responsibility for every sentence in that book as a co-author, but that wouldn’t be zero-level responsibility (responsibility for having done X).
“Right, but how can I be responsible for having said that there are mortals if I say that Socrates is a mortal? I did not say that there are mortals. I did not perform that action.”
We did talk about this, remember. Suppose that the water boils by itself, miraculously, when I put the teabag in it. I did not boil the water, yet I am, in a sense, responsible for having boiled it, if boiling the water is a necessary condition for making the tea. Opining that there are mortals is, in a sense, a necessary condition for opining that Socrates is one, but we can miraculously do the first without doing the second.
“It seemed far-fetched when you first said this and it still seems far-fetched now.”
Ok. The problems seems to be that responsibility entailment for regular actions seems to be rooted in causal dependencies, while logical implication, conceived as a relation between sentences, is not. Also, there is something strange with talking about logical relations between speech acts, since they, conceived as communicative actions, are actions, that is events. The event of uttering “Socrates is mortal” can, of course, occur without the occurrence of the event of uttering “There are mortals”. Generally speaking, any utterance of a sentence can occur without the utterance of the sentences logically implied by that sentence. So it would seem that there is only one logical implication according to this view: X, therefore X.
Now, saying that it is ‘as if’, when one utters “Socrates is mortal”, one utters “There are mortals and one of them is Socrates” does not help much. Why should it be so? Also, how much can we include in the content of an utterance to save all the possible inferences?
Things would change if, instead of talking of “having done B”, we were to talk of “doing B”. Thus, we could change the definition of responsibility entailment by saying: “By being responsible for doing A one is responsible for doing B.”
At this point we could say that responsibility entailment is a relation between responsibilities, not actions. Suppose I did not utter A, but I assume responsibility for uttering it. On could assume such a responsibility for various reasons (somebody else could have uttered A in my name etc.). I am responsible for uttering B.
“But are you still talking about zero-level responsibility now? You seem to conclude that you have a duty to say B if prompted about it. This is not the responsibility for the performance of an action anymore.”
Ok. Let us distinguish between:
a) being responsible for uttering A
b) being responsible for asserting A
When I say “I have uttered A” I am describing my communicative action, so to speak, while when saying “I have asserted A” I am describing the speech act I have performed. One can utter A without asserting A, of course. Why not say that one can assert A without uttering A?
“You could assert A by writing it down, of course.”
This is not of much help. What if I hire a speaker to utter A for me? What if I utter it in a different language. Should you refrain to render the content of my assertion in English? (Am I making an assertion only if I utter “I assert that”?)
“Now you seem to make speech acts into something abstract. They resemble type-sentences conceived as linguistic objects. But one cannot perform abstract actions.”
Oh, well. Can one not open a ceremony? People perform ‘abstract actions’ all the time. They befriend each other, grant honors to each other, divorce each other and so on. Ontology will only mislead us here.
Let us call “George” the particular utterance of some particular words. An individual event, if you wish. By performing “George” one can assert that Socrates is mortal. What are we talking about when we say “he asserted that Socrates is mortal”? The performance of George, of course. That does not mean that we want to say that the content of his assertion were those particular words. If it was so, nobody could be making the same assertion.
“Making an assertion”, however, does not work like “making a house”. Neither does it work like “making a hole in the wall” or “making a shadow”. Nor like “making a fuss” or “making a full of yourself”. It rather works like “making a pact” (or “taking a step”). We do not actually believe that people making assertions do actually perform some sort of abstract actions. There is no ontological commitment in what we say about what they do.
“But something does effectively happen when a person takes a step.”
Sure, but it’s not “the taking of a step”. Something also happens when two people (or even two institutions) make a pact.
“Ok, but you cannot have your cake and eat it too. If in talking about U’s assertion that A we talk about what has effectively happened, the event that U has uttered some particular words, for which U is responsible, we cannot say that U is also responsible for asserting B, since in that case nothing did effectively happen. How could one be zero-level responsible for an action when nothing did happen?”
I am responsible for asserting that Socrates is mortal. Something did happen. I might have uttered some words or my personal speaker did a.s.o. Now, by responsibility entailment, I am responsible for asserting that there are mortals. It is not like nothing did happen. Something did. The same thing I was talking about when I said “I might have uttered some words or my personal speaker did a.s.o.”.
“But when you usually assert that there are mortals something different happens. The utterance of other words, to be more precise. Do you want to say that someone can assert that there are mortals by uttering the words ‘Socrates is mortal.’? And how is this different from trying to fit in the content of one’s utterance everything which is logically following from it?”
Ok. Then let us try something else. Instead of zero-level responsibility (responsibility for having done A), let us talk of “responsibility for doing A” as something different, such that one can be responsible for doing A without doing A. I would like to distinguish this, however, from the duty-like responsibility to do A, since we do not want to say that the performance of A is entailed by the promise to do A.
“But the responsibility to do A is entailed by the promise to do A. When you have reached that conclusion you were talking about non-entailment between the actual performance of a speech act (promising to do A) and the actual performance of A, but now the situation has changed.”
We are faced with two choices:
i) Saying that action A responsibility entails action B IFF: if one is duty-responsible to do A, then one is duty-responsible to do B.
ii) Saying that A responsibility entails B IFF: if one is zero-level responsible for doing A, then one is duty-responsible to do B.
However, i) does not hold, since saying that Socrates is mortal makes you responsible for saying that there are mortals even if you did not have any duty to say that Socrates is mortal. ii) does not hold either, because you can be zero-level responsible for saying that Socrates is mortal without having any duty to say that there are mortals (in the way in which you might have a duty to tell your child at some point that she/he is not immortal).
“In any case, isn’t it pointless to look for a definitions which you can only give in the form of a conditional? Why not just say that ‘B logically follows from A IFF if A, then B’?”
The metalanguage of my definition (which can include terms like “necessary condition”) is not at stake here.
“But it is, since boiling water is not a necessary condition for making tea. Not in the sense in which we speak about logical necessity, in any case. Remember that not being green all over is also not a necessary condition for an object being red all over.”
I think we can move past that concept of strong logical necessity. We do not need to boil water in order to make instant tea and it is accidental that we decided to call that drink “instant tea” and not something else. The same can be said, perhaps, about not changing the reference of a proper name with each of its occurrences, such that it would be accidental that we take ‘John is John’ to be always true (see Searle, Proper Names). Let us put this aside for now.
To resume our main line of thought, let us call one kind of “responsibility for doing A” which is not zero-level responsibility “proxy responsibility”. When my personal speaker speaks for me I am proxy-responsible for what she says. When my child hits your child, I am proxy responsible for what my child did. Now, the substitution which the word “proxy” is about does not have to be only the substitution of the agent. It could be the substitution of the content of an action as well. Also, such a substitution might not only regard the substitution of words figuring in the description under which the action is considered with synonymous words. When Anscombe is discussing the ‘pushing the lever – acting the pomp – providing water to the hose – etc.’ case, she seems to be talking about proxy responsibility, in this sense.
There is, still, another sense in which we can talk about “being responsible for doing A”. This is, as in the case of proxy responsibility, the case in which you can hold me responsible for A even if I am not zero-level responsible for A, but no substitution is needed in order to hold me responsible for A.
Why is that so? Perhaps, it is simply that I had somehow assumed responsibility for A. By saying that Socrates is mortal I have assumed responsibility for asserting that there are mortals. Also, if I started by talking about Socrates as a non-fictional character, I have assumed the responsibility to talk about him in the same way in regular contexts, no matter what I say. If I do not do that, you can hold me responsible (“as if” I have given you assurance you that Socrates is a real person). If I say that there are no mortals after having said that Socrates is mortal, you might hold me responsible.
“Ok, but I will hold you responsible for having said that Socrates is mortal, not for having said that there are mortals. Also, I do not see how can this account of yours distinguish between what an assertion entails and what it assumes. Also, if you start an action series, in some context I might hold you responsible for not finishing it, but you do not want to say that the first actions in the series entail the following actions.”
If I start making a soup and then I deliberately spoil it, you might hold me responsible for what I did. And I could provide you with a justification for what I did by pointing out that each of the two actions was part of different series of actions and not of a single series. Alternatively, I could admit that spoiling the soup was wrong (and not just ‘a mistake’) or that starting the soup was wrong (in this case, if I just said that it was ‘a mistake’, you could insist that I did not have to spoil it). [this example is not very good, I think]
[I feel like I am hitting a wall again and again; there has to be a door somewhere near, and yet I cannot find it.]
In any case, when you hold me responsible for finishing the action series which I had begun, you think I have the duty to do it. So it’s duty responsibility which you are talking about (and not my “close to zero-level, close to proxy responsibility”). Also, if I say that George quit smoking and then I add that he never was a smoker, you will hold me responsible either for saying that George quit smoking (“why did you say that, what did you mean by it etc.”) or for saying that he never was a smoker (“what did you mean etc.”), but not for contradicting myself. On the other hand, if I said that Socrates is mortal and I add that there are no mortals, you will hold me responsible for contradicting myself. There is no direct logical contradiction between my two assertions, so you must consider me responsible for asserting that there are mortals and there are no mortals in order to do so.
Now, suppose you could know that I have made the tea without knowing whether I did boil the water or not. You will still hold me responsible for boiling the water. That is, you will talk to me as if I did that.
“It’s your ‘miracle’ argument all over again.”
No, it isn’t.
“So what do you want to say, after all. Logical entailment should be conceived as a relation between actions, because you can talk of it in terms of the ‘close to zero-level, close to proxy responsibility’ which one has in virtue of doing (including saying) something?”
Indeed, it has to be so. Otherwise, we are going to have an abstract logical space filled only with sentences (or propositions) completely cut from all that is happening in our lives.
“And why must logic have something to do with our lives?”
Well, if you do not believe that logic has anything to do with reason, it must not. But cannot see how could reason be unrelated to our lives.
“Aren’t causal relations (or natural laws) the expression of reason in our lives, lives which, in their turn, are part of nature itself?”
Oh, please don’t start again with this.
“Ok, but I do not understand your talk about reason.”
What don’t you understand?
“To begin with, I think reason guides me in my actions when I am making decisions or when I optimize my actions. I do not feel like I am guided by reason all the time, whenever I do something. Aren’t you just a disguised intellectualist?”
I am not talking about reason as some sort of faculty, the workings of which are available to us by introspection. Just think about it. When an internalist says that a belief must be justified in order to represent knowledge, she does not talk about an inner psychological act of justification. It is not reasonable to start making a soup and then to deliberately do something to spoil the soup. My talk about opposing actions is an attempt to conceptualize such cases in such a way that when we say ‘it is not reasonable’ we do not change our talk too much. We say it in pretty much the same way as when we say that ‘it is not reasonable to contradict yourself’. There is no need to talk about psychological processes or of some rational faculty here.
It is the same with the relation of inference (if you dislike talk about logical implication). We say that from ‘Socrates is mortal’ it follows by reason that ‘There are mortals’. Other examples can be provided if needed. Now, if justification presupposes inferences and its starting point are the actions we perform in our environment, then some beliefs must ‘follow by reason’ from such actions.
So if I perform the action of sitting myself in a chair (under this description), it must follow by reason from my action that there is a chair close to me. I do not want to say that the action causes my belief (although it can be so), since I would not be talking about justification in that case. So I want to talk about my action and that belief in the ‘logical space of reason’, since that is the space needed for justification (among other things). However, conceiving the belief as a propositional attitude, that is, a mythological hybrid between a psychological intentional state (believing that…) and an abstract object (p) does not seem helpful here. So I prefer to talk about ‘acts of opining that’ instead of beliefs. My belief that there is a chair close to me is empirically justified. For me, this means that my act of opining that there is a chair close to me (or some other act of attributing that belief to myself) ‘follows by reason’ from my empirical actions (sitting myself in the chair, looking at it, resting my legs on it etc.).
In a similar way, I think we want to hold that people can act reasonably based on what they know and aim. It does not matter how often they do that. Their actions can be justified or not, given what they know, their purposes and perhaps their norms. This can be conceived as the other side of justification, leading from what we say to what we do. If I want to make a tea and I know that in order to make a tea one has to boil water, then my action of boiling the water is justified. That is, it follows from my (justified true) belief that one cannot make a tea without boiling water and my intention to make a tea. Again, I prefer to talk about the action of making a tea and the action of boiling water only, in such a case, and say that the second action follows by reason from the first. In some other cases, though, I am prepared to say that a non-communicative actions follows by reason from a communicative one. My actions performed as a parent towards my children are justified by my acceptance of rules with respect to what I ought to do as a parent. My acceptance of those rules comes down to the speech acts I would perform if asked to justify those actions.
[I should say more about this]
Let me work a bit on the ‘personal-intentional-actional-responsibility’ vocabulary.
X (a person) does A (an action) IFF:
(i) A occurs; here we already have some problems events do occur, but perhaps we want to say of actions that they are performed (by somebody); nevertheless, we can include the performance of action A in the description of event E: X’s body moved such that he opened the door; but now, if the second condition is that X assumes zero-level responsibility for A, the problem is: why not E? Why say your responsible for opening the door and not for your body movement which led to the opening of the door? Perhaps, because you do not want to assume responsibility for all your body movements, described at some physical level. Not all of them are basic actions. That is, not all of them are distinct steps in some series of actions (version: not all of them can be part of a single step series of actions). In any case, we do not say “I have moved my body such as to open the door”, but “I have opened the door”. Suppose one could open the door by thinking to do it. A device reading the neurophysiological processes in my brain could wirelessly activate a mechanism embeded in the door etc. Should I say that I have done something different from opening the door, only because I am opening it without moving my body?
One can learn to do A in more than one way (incidentally, can one do A without having learned to do A?). Some causal inwards of our actions are not under our control, but this does not make us say we are only responsible for the parts under our control (see ‘shooting Bill’ vs. ‘pulling the trigger on Bill”).
Also, is this a necessary condition? How couldn’t it be? Why say you did A if A did not occur? Of course, I am not looking for a motive here. Not even for a reason for saying it. People can say unreasonable things. The question should have been: ‘Why accept that somebody did A if A did not occur?”. Suppose some series regularly including A has concluded without the occurrence of A. You prepare a tea using hot water from your water dispenser and say that you have boiled (or heated) the water. You wouldn’t have said that if the water dispenser did not belong to you.
It could be replied that even if we say that the person did heat the water, it is not as if the respective event (heating of the water) did not occur, since it did. The causal chain linking the person to that event is perhaps longer (leading from the presently heated water to the acquisition of the water dispenser etc.), but it’s still there.
What if X puts a rock into a cup, covers it with a towel and when she uncovers it the cup is full of hot tea? Suppose something like this happened. We’ll perhaps say that X has turned a rock into tea, but are we supposed to say that X has boiled some water as well? I do not think so. But even if we did, we would say, perhaps, that X has turned the rock into water and tea leaves, has boiled the water and made the tea leaves disappear afterwards. So if X did boil the water, then the boiling of the water must have occurred. The problem is that we cannot explain how did it occur, not that we say it didn’t. I do not think one could ever say ‘X did A, but A did not occur’.
(ii) X assumes zero-level responsibility for A – that is to say, if we wanted to criticize (or praise) someone for A (version: to understand the occurrence of A in terms of someone’s personal reasons for A), X would accept to be our interlocutor; the simple version is that X would answer the question ‘who did A?’ by saying ‘it was me’ or the like.
“What if X does not want to recognize that she did A? Shouldn’t we say that she did it?”
The condition is not that X assumes responsibility for A publicly. But if X denied having done A even in her private diary, we’ll be left only with the causal connection between X and the event A. We could say that X is the cause of A, but not that X did A.
“What about X’s intentions?”
My intention that A occurs does not change anything if I do not assume zero-level responsibility for A. Suppose I intend the door to open and a machine reading my brain opens the door, but I know nothing about the existence of the machine. My intention to open the door was the cause of the door opening, but for me it was a coincidence that the door has opened then. If I learned about the existence of the machine, I would still say about this first case that the opening of the door was not my doing. After all, I did not intend the door to open in order to cause it to open (I did not intend to perform the action of opening the door), since I did not know that I could open it by simply intending it to open.
Of course, it could be said that by ‘intending the door to open’ in the previous paragraph I have meant ‘wanting the door to open’. This amounts to saying that intentions (or at least the relevant intentions) must have actions as their content, but then the definition would be circular.
“Isn’t this also the case with your zero-level responsibility?”
No, since one can assume zero-level responsibility for the occurrence of an event. For instance, a child can assume zero-level responsibility for a dice falling on a certain number.
“Right, but the child believes that she ‘had thrown a six’. So what she assumes responsibility for is still an action from her point of view. It is only us who talk about the ‘dice falling on six’ as of an event.”
What if my son did A and I assume responsibility for A? I am not assuming responsibility for his action (his breaking of the neighbor’s window, say), since it is his action and one person cannot assume responsibility for another person’s actions, but I can say that I am responsible, in a sense, for the breaking of the neighbor’s window.
“You can say whatever you want, but you shouldn’t say that you are zero-level responsible for that, since that would mean that you did broke the neighbor’s window and not your son. In any case, if you would manage to conceive zero-level responsibility such that one could be responsible for the event A without being responsible for the action A, then it will follow from this that conditions (i) and (ii) are not sufficient, since one can fulfill them without having performed the action A. So you’ll have to add another condition and so on, ad infinitum.”
Ok. Perhaps I should have said that one can assume responsibility for the occurrence of the event A whenever one performs the action A. When I boil some water I assume zero-level responsibility for the water boiling.
“But also for boiling the water, right?”
Perhaps we could say that zero-level responsibility assumptions can only concern events.
“Ok, but now you have to add another condition to your definitions, since you do not want to say that the child did throw the dice on six.”
We can leave the child aside and imagine the case of a person mistakenly believing that she killed somebody. Laura intends to kill George, buys what she believes to be a poison, puts it in George’s glass, George drinks it and dies of poisoning due to unrelated causes (unknown to Laura, somebody else put real poison in the same glass etc.). So we can add:
(iii) X is justified in assuming zero-level responsibility for (the event) A.
Now, the easiest way to talk about this seems the be saying that X is justified in assuming zero-level responsibility for A IFF has caused A, but that would be wrong.
Take Laura’s case. George was poisoned and she thinks she has poisoned him, since she put some powder into his glass, which she mistakenly believed to be poison. The poison was put by someone else into the wine bottle and she poured wine into George’s glass, so she caused George to be poisoned, but we should not simply say that she has poisoned him. In any case, she is not justified in assuming zero-level responsibility for George’s poisoning, even if she caused it. Justification cannot simply be reduced to causation.
So what is it to be justified in assuming zero-level responsibility for an event? Aren’t we talking about justification for the belief “I did A”, after all? I do not think so. First, there is a difference between:
(a) I did (action) A.
(b) I am (zero-level) responsible for event (A).
Even if one couldn’t believe (a) without believing (b), one could believe (b) without believing (a).
“Wait a bit. In this case, even conditions (i)-(iii) are not going to be sufficient, since one could believe (b) and even be justified in believing (b) without believing (a).”
Suppose I can move objects ‘with my mind’. Now, suppose I think my power works via some ‘divine cooperation’. I might believe that I am zero-level responsible for moving a ball and be justified in believing so, but refuse to believe that I have moved the ball, since I mistakenly believe that I did not do it alone.
In any case, one could assume zero-level responsibility for A without believing that one is zero-level responsible for A. This was my second point. Beliefs conceived as propositional attitudes, i.e. mental states with a propositional content are not necessary for assumption of zero-level responsibility. I assume zero-level responsibility for writing every single letter from this text (since I did not use an auto-correct / auto-complete software), but I did not have all the psychological states of ‘believing that I am responsible for the appearance of a ‘W’ on the screen, believing that I am responsible for the appearance of an ‘h’ on the screen a.s.o.
One could talk about psychological states which are not accessible by introspection, of course, but I think we can avoid that.
Let my propose, then, that X is justified to assume zero-level responsibility for (event) A IFF A is part of series S and all the steps preceding A in S were performed successfully (version: correctly) by X. That is, if Laura did buy poison and not a harmless powder, she would have been justified in considering herself responsible for poisoning George.
“What if somebody removed George’s glass and replaced it with a different glass of poisoned wine?”
She would still have been responsible for _putting poison in George’s glass_ and this is what I have meant by ‘poisoning George’.
“But you were talking about George being poisoned, not about his wine being poisoned.”
Ok. Perhaps Laura is not justified in this case to assume responsibility for George’s death by poisoning. That event, however is at the end of a longer series of actions, which include seeing that George drinks the glass of wine with your poison. Laura did not do that successfully. If she did, she would have stopped the other person from replacing George’s glass.
“Ok. Suppose, then, that the recipe for some magic potion included ‘leave the ingredients unattended for half an hour’ as a necessary step. During that half an hour someone replaced the ingredients with the final product. Shouldn’t we say that even if the person trying to make the magic potion was justified in assuming responsibility for making the potion (since she performed all the steps correctly/successfully), she nevertheless did not make the respective potion?”
For your example to work you need a recipe stipulating that one should refrain from preventing another person from interfering with the ingredients as a necessary part of making the potion. If that was the actual recipe, perhaps we should say that the witch did make the potion.
I suppose you are looking for a case of the following type:
– X performs all the steps from series S correctly, up to A.
– X assumes zero-level responsibility for A.
– X does not cause A to occur.
Let’s give this another try. Suppose I want to open an automatic door. In order to do so, I have to walk towards it, since the mechanism opening the door activates only if one is closer than a certain distance from it. I do this successfully and the door opens in front of me, but not because my presence in front of the door has activated the mechanism. The mechanism is broken, so someone manually opens the door every minute and it just happened that I was around at the right moment. I could not have opened the door, since the other person did it. There is no causal connection between my presence in front of the door and the opening of the door.
The causal relation missing from this story is not that between myself and the door opening, but that between my walking in front of the door and the door opening. But even if the causal relation was in place, that would not solve the problem. Suppose that the other person does not open the door every minute, but watches the place on a surveillance camera and opens the door when she sees somebody approaching it. In this scenario, my presence in front of the door did cause the door to open, but I did not open the door.
These cases, however, might not call for a revision of condition (iii), but rather for the addition of another condition.
“Before you get to that, I want to point out that your understanding of ‘being justified to assume zero-level responsibility for (the event) A’ involves the concept of ‘X performing action A’, since you talk about X having performed the necessary steps in series S correctly, and those steps are actions.”
You’re right, but this only means that some coherentist / holist conception of justification is at work here.
“Ok. Do go on, then.”
Well, I suppose that our last cases (with the automatic door) do resemble Gettier’s counterexamples, in a way. We need to bridge the gap between X’s justified assumption of zero-level responsibility for (the event) A and the occurrence of A, such that the satisfaction of those conditions would amount to X having done A.
“Sorry for the interruption, but what if A is the only step in series S? It would appear that one is justified in assuming zero-level responsibility for A unconditionally. How could this be?”
Either A can be a step in a series S*, but it is the single step of the series S, which I perform right now, or it can only be the single step in series S and no part of any other series. Now, the second case is not possible. Any action can be a step in a series which includes other steps. It could be, however, the first step. Again, A could be such that it is not the first step in some series S*, or it could be such that it is the first step in all the series in which it appears. This is doubtful, but I will stick to your case.
Let us imagine a person who believes she can move the leaves in a tree ‘with the power of mind’ and that in doing so she performs a single action which requires no preparations. Is she right in believing that? I mean, she has at least to watch the leaves she wants to move. (perhaps ‘watching X’ is the kind of empirical action which figures as the first step in many series – although not in all series – but it could also figure in a series as a further step). She has to do this in order to identify the leaves she is trying to move, but even if she only intends to move ‘some leaves’, she has to make sure that they are not already moving. And even if she makes sure of that in some other way, in order to do that she has to perform other actions. It could be objected that those are not part of the series called ‘moving some leaves’, but then she has at least to determine the leaves’ position, in order to see whether they will move or not. But suppose that she did not care to check this out. If so, how could she be justified in assuming zero-level responsibility for A? You cannot be zero-level responsible for A if A did not happen. We agreed to that. So it seems that justification for assuming zero-level responsibility for A has to include those empirical actions by which one makes sure that A did happen. However, since A is part of a series S, it seems that those actions are also part of the series S, so perhaps we only need to modify (iii) a little bit, by saying that the steps necessary for the continuation of S ‘from A on’ must be performed successfully.
This does not end here, of course. I suppose you would be thinking of a case in which the leaves do accidentally move. So X watches the leaves moving and assumes zero-level responsibility for having moved them and this is all that happens. I suppose you want me to say that X is not justified to assume zero-level responsibility for having moved the leaves because she does not know the real cause of the leaves’ movement (whatever that is), or at least she does not know that she is not the cause of the leaves’ movement.
Let’s take another case. The doctor hits X’s knee with a hammer and X’s leg moves. In X’s view, she is responsible for her leg moving. You say that my conception makes it possible that she is justified in assuming responsibility for her leg’s movement. Perhaps it should be like this. If X thinks that the doctor is cooperating with her in moving her leg and the cooperation consists in the doctor providing the cause of the movement while she provides the intention to move her leg, perhaps this is ok.
Were I to discover that all the movements of my body were produced by some external cause, I would not cease to take responsibility for my actions, which means, in turn, that I would not cease to assume zero-level responsibility for those movements. I might say “I intended my body to move in those ways”, but that wouldn’t mean that I believe my intentions, conceived as psychological acts, to have also caused those movements.
So perhaps in the ‘automatic door’ examples, the only problematic cases are those in which it could not be said that I have opened the door because somebody else did it (the doctor is indeed somebody else, but X does not believes the he ‘moved her leg’; he only ’caused her leg to move’). As long as the door opens as I intend it to open, the mechanism causing it to open does not matter. So perhaps our last condition should be:
(iv) There is no Y, such that Y does A.
“So, in conclusion, what you want to say is that in order to conceive an event as my action, the event in case must either occur in a series of events which I already recognize as actions performed successfully by myself, or, if no other events recognized as actions are necessary for its occurrence (but not in any causal sense), it must not be claimed by somebody else as that person’s action?”
I am not talking about claims here.
“I know, but I cannot stop from thinking of cases in which two equally mistaken people claim to have moved some leaves in a tree with the power of their minds. It seems that the first to make his claim wins, according to your view.”
This is a weird case, indeed, but I suspect that its weirdness comes from your expectations. You expect us to establish the cause of the leaves’ movement, while I am not interested in finding the cause, but the person responsible. Think of two people pushing a large enough button at the same time. Now, if neither of them wants to assume collective responsibility for pushing the button and each of them wants to be entirely responsible for pushing it, we are in the same situation. We can say that the first one claiming responsibility for pushing the button wins and it’s going to be equally weird. Alternatively, we could say that both did, even if they did not do it together, on pain of giving up the principle that two persons cannot do the same particular action non-collectively. These are, however, extreme cases.
“You did seem willing to discuss some pretty extreme cases up to this point.”
Then maybe we will discuss your ‘multiple claimants’ cases some other time as well. For now, I think that at least a sketch of my view of actions is in place. I do not think that we need to talk about causal connections in order to conceive some events as our actions. Instead of causal connections it is enough to talk about rational entailment between actions within action series. I am justified in assuming zero-level responsibility for A, because if A was my action, it would have rational entailment relations with other actions performed successfully by myself.
“Ok, but you have conceived those relations as ‘close-to-zero-level-close-to-proxy’ responsibility generating relations. Aren’t you moving in a circle after all?”
I do not think so. In terms of responsibility, I have distinguished at least five things:
(1) X is zero-level responsible for the (event) A
Which means that:
– A did occur
– X is assuming zero-level responsibility for (the event) A
– X is justified in assuming zero-level responsibility for (the event) A
[if X is zero-level responsible for the (event) A, and nobody else is zero-level responsible (not co-responsible) for A, then X did A]
(2) X is ‘close-to-zero-level-close-to-proxy’ responsible for (action) A.
Which means (in short) that:
– it is as if X did A (if by doing A, one offends me and X is ‘close-to-zero-level-close-to-proxy’ responsible for A, then X offended me);
(3) X is account-giving responsible for (action) A [towards Y]
– Y can ask X for reasons for doing A
(4) X is (consequence) responsible for A towards Y
– X did A
– Y was either the addressee of A or the one mainly affected by A
– X has to do something with respect to A’s consequences for Y
(5) X is duty-responsible for (action) A [towards Y]
– there is a rule according to which X ought to do A
– Y may hold X responsible for not having done A
Now, a more simple view might emerge if we noticed the most important difference between the extremes of this list – (1) and (5). Namely, in (5) X is responsible with respect to a possible action, while in (1) X is responsible for an occurring event. Whether the event A did occur or not is essential in (1), but of no relevance for (5).
(4) and (3) can be conceived in terms of duty-responsibility, since X has the duty-responsibility to provide reasons for doing A to Y at (3) and to perform other actions with respect to the consequences of A for Y at (4).
(2) is perhaps also conceivable in terms of duty-responsibility, although it is not about the duty to provide reasons for doing A or manage the consequences of doing A (since you might not have done A), but it could be about the duty to provide reasons and manage consequences as if you did A.
The alleged circularity, now, comes from the following:
The content of duty-responsibility can only be an action. So X can have a duty to do A only if A is an action.
In order for A to be a (possible) action of X, according to the above definition, A must be an unclaimed event which X is justified to assume zero-level responsibility for.
X is justified to assume zero-level responsibility for (the event) A only if X has performed the other steps in the series S, except those from A on, successfully.
Those other steps are excepted because they are not necessary conditions for doing A.
The necessary steps (which are not excluded) are, in a sense, rationally entailed by A. That is, they would be rationally entailed by A if A was an action.
Suppose B was such a step. Then, by performing A X would be duty-responsible as if X did B.
But now we can say what means to be duty-responsible for doing A only if we say what it is to be duty-responsible for doing B. Even if we accept this approach as a valid holistic approach, it seems that we cannot reduce duty-responsibility to zero-level responsibility.
But why shouldn’t this be ok? After all, this goes hand in hand with my initial intuitions:
– that rational entailment depends on duty-responsibility; that is, without a concept of duty-responsibility one cannot have a concept of logical consequence.
– that without embedding events into the space of logical consequences / rational entailment, one cannot conceive some events as one’s own actions.
“Stop here. Can’t you see that you’re running in circles now? There is one line left: ‘Without conceiving events as actions one cannot have the concept of any duty-responsibility’. All is pointless.”
No, it isn’t. I suppose that these concepts have grown up together. But I will have to think more about this.
“Even if your not running in circles, you do not seem to get anywhere. Having traced the normativity of language to our practices, which you conceive as being composed of actions, you claim that the normativity of our actions resides in the commitments we undertake (or the responsibilities we assume) when we perform them. You talk, in fact, about an unchallenged justified assumption of responsibility for the event A as being equivalent with performing the action A. But the simple assumption of responsibility (justified or not) cannot be an act (since if it was, you’ll have an infinite regress here). Thus, it has to be reducible to a natural event, but all normativity will disappear at that point. Talk about assumption of responsibility seems to be meant to replace talk about intentions (conceived as psychological states). But this does not seem to make it better.”
First, I do not want to avoid talk about intentions at any cost. I only want to distinguish between talking about intentions while undertaking an ontological commitment (“intentions are something”) and talk about intentions with no ontological commitment (“what were your intentions” simply means “why did you do it?” – this asks for a justification and not for introspection).
Assumption of responsibility is a necessary condition for asking for justifications. One can ask X for her justifications for doing A only if A assumes responsibility for A. However, the assumption of responsibility for A does not have to be an event having happened before on asks “why did you do A?”. If X replies to that question by providing her reasons for doing A (reasons which she does not have to have thought of before having done A), then she has assumed responsibility for A.
The assumption of responsibility is not something. If it was, your critique would apply, but that is not the case.
“But one can explicitly assume responsibility for A, by performing a speech act, right?”
Sure but the speech act could have the form of “I accept questions about my reasons for doing A”. In any case, the explicit performance of such a speech act is not necessary, just as saying “I accept to receive invitations” is not necessary in order to receive invitations.
“Then, if I ask you about your reasons for blacking out the sun after a solar eclipse, does this make you responsible for the eclipse?”
If I answer your question, I make myself responsible for it (in front of you, at least). But I can reject the question, of course.
“Then one assumes responsibility for A if one does not reject questions asking for her/his reasons for A, isn’t it?”
I suppose I could agree, since not rejecting a question is not an act (in contrast, from instance, to “refraining from rejecting a question”). “Asking a question”, in this sense, seems to resemble “betting” (or “making a promise”). All can mean either the proposal (to answer, accept the bet or the promise), or the successful result of the proposal.
“You seem to overlook something. Since you make “not rejecting the question asking for your reasons for A” into the sufficient (and perhaps also necessary) condition for assuming (zero-level) responsibility for the event A (we ignore here the further complication that one cannot provide her reasons for the occurrence of an event) and the assumption of responsibility is a necessary condition for the performance of an action, you make the performance of actions dependent on linguistic abilities. Surely you do not want to say that people started to perform actions only after they were able to ask for reasons and provide reasons.”
Why not? I did not say that asking for reasons and providing reasons can be done only by using a language. Since a person can provide her reasons for doing A by doing B (which is not a verbal action), why couldn’t a person request a reason for doing A by doing B?
In the first case, the reason could be provided by performing B, which A was a necessary condition for. For instance, if you ask for my reasons for boiling the water I could tell you that I have boiled the water in order to prepare a tea, but I could also show you that I am making tea by pouring the water in a cup and putting a teabag in it. It might be that “simply doing A” and “doing A to show someone what you were trying to do” do not mean the same thing, but you do not need language mastery in order to do something to show someone what you were trying to do. A young child could do that before learning a language (my younger son brings me his shows to show me that he wants / intends to go out, for instance). Now, if such actions can be considered reason providing nonverbal actions, then the reason requiring nonverbal actions should be the actions which can prompt the performance of such actions. For instance, in order to find out your reasons for doing A I could attempt to oppose A (by performing actions which are opposed to A – either preventing you from doing A, or undoing A etc.) at first, but then let you do it (or even encourage you to do it) while ostentatiously watching you.
“You still talk of communicative actions. So even if one does not need to master a language in order to act, you seem to claim that one needs to be able to communicate in order to act.”
Well, a rudimentary form of communication seems to be required, indeed, but I think we can accept that even some evolved animals can communicate in this way.
“So they can assume a responsibility for what they do as well?”
They can act. We are talking about zero-level responsibility, remember? Being able to act does not necessarily make you a moral agent, if that was your worry.
“Ok, I will not insist with this for now. I have, however, another worry. What are you going to say of the last action in a series? One cannot provide reasons for that by doing what that action was enabling her to do. What about the last thing one does before one dies. It seems that the last thing you do cannot be your action, according to your view.”
Were I asked to provide my reasons for it, I would have accepted that request (one can accept the request without being able to complete it, mind you; thus, one can assume zero-level responsibility even for gratuitous actions). Now, we tend to treat such a counterfactual as if it has an ontological commitment, if not to some events or mental states having had occurred, at least to some psychological dispositions. We also regard regular dispositional terms (like “soluble in water”) as being reducible to some physical properties, so we assume that psychological dispositions are also reducible to such properties. Reasons-accountability, I want to say, should not be seen like that. After all, it is not the disposition to do something. But if you want to talk about “the disposition not to reject a request to provide your reasons for A”, so be it. If I had that disposition when I did that last action (either in a series, or in my life), then I count as having performed it.
To escape the objection that someone cannot be said to provide reasons for the occurrence of an event I could, perhaps, say that zero-level responsibility is the disposition not to reject requests for your reasons for having caused the event A. Of course, one does not need to have the concept a causation in order to have such a disposition. The concept is needed only in order to talk about this disposition – or to make it explicit, as Brandom might put it.
“There is still a problem left. Namely, you seem to say now that the ostentatious performance of B can count as providing a reason for doing A if A is a necessary condition for B, but previously you have said that if A is a necessary condition for B, then A would rationally entail B.”
You must have got me wrong. What I was trying to say was that if A was a necessary condition for B, then B would rationally entail A.
“Well, you did say: ‘X is justified to assume zero-level responsibility for (the event) A only if X has performed the other steps in the series S, except those from A on, successfully.’ But now you say that the performance precisely of a step ‘from A on’ in the series S can work as a reason for doing A.”
Indeed it can, but this does not mean that the performance of such a step is necessary for assuming zero-level responsibility for A. Let me recap:
– X assumes zero-level responsibility for A only if: if X was required to provide reasons for causing A, X would not reject the requirement (but X could accept the requirement and say that she cannot provide reasons; also, X could provide reasons by saying something or by doing something; in this last case, she would perhaps do B, such that A was a necessary condition for B).
– X is justified to assume zero-level responsibility for A only if: if A was an action, A would rationally entail other actions successfully performed by X (version: X did perform the actions rationally entailed by A) and: no one other than A is zero-level responsible (non-cooperatively) for A (this could perhaps be discussed further, but I am not saying that these conditions are sufficient; X is justified to assume zero-level responsibility for the first action ever performed by X only if the event in case could be an action but nobody else is zero-level responsible for it);
– If X assumes zero-level responsibility for A and is justified to assume zero-level responsibility for A and A did occur, then X is zero-level responsible for A.
– If X is zero-level responsible for A, then X did A.
– There is a rule saying that X ought to refrain from doing A / ought to do A only if: there is a practice such that by taking part to it you commit yourself to refrain from doing A / doing A and: X takes part in that practice.
(one commits herself to doing A (in the appropriate circumstances) explicitly only if one promises to do A (in the appropriate circumstance) or performs other speech acts to the same effect; the same with refraining from doing A / not doing A; one implicitly commits herself to doing A if: if one was asked to explicitly make that commitment, one would do it; one takes part in a practice if: if the person was to undertake the explicit commitments specific to that practice, the other practitioners would accept her promises; this is not the best treatment of this issue, but I have no better ideas for now)
– X is duty/consequence responsible for A only if: X did/did not do A and: there are rules saying that X ought to have refrained herself from doing A / ought to have done A.
– X is morally responsible for A only if: X is duty/consequence responsible for A and: the rules saying that X ought to have refrained herself from doing A / ought to have done are moral rules (R is a moral rule, in short, only if it concerns how persons ought to treat each other – infringement of such a rule coincides with treating someone as something less than a person)
[as a side note: I can have duty-responsibilities towards a pet, which means that it is not morally wrong not to fulfill my responsibilities towards my pet, but this does not mean it is not wrong]
So what is it to recognize X as a person? Perhaps it is to recognize that X can have duty-responsibilities. Treating X as a person, then, would mean not interfering with X’s duty-responsibilities. For instance, if I kill X or steal from X, I prevent X from keeping her duty-responsibilities (or from assuming certain duty-responsibilities towards what X has already done).
There are several gaps in this account, but this is just a sketch, of course. I am not especially interested in moral responsibilities, in any case. That is just a species of the larger genus of duty-responsibility which I am concerned with.
Of course, I do not want to say that ‘It is possible that X has duty-responsibilities’ is a sufficient condition for ‘X is a person’. It might be possible to turn any cat into a being having duty-responsibilities (the cat should be able to make promises, but we can think that it could do so). This, however, does not make cats persons. Modal and counterfactual talk should be accounted for in order to say more about this.
Also, one might wonder about the natural roots of commitment. A rudimentary kind of commitment seems to be needed even for performing actions (that is, for assuming zero-level responsibility for an event), since ‘not rejecting reason-providing questions’ can also be called ‘committing yourself to answering reason-providing questions’. Of course, I can turn this around and say that commitment is not something. It is not, for instance, a special act by which one accepts a certain practice (that of asking for reasons and providing reasons, for instance). It can be, simply, the lack of a rejection of that practice. The practice in itself could have grown from our needs to adapt to our environment, to better our chances of survival etc.
“This is awkward. You want to develop this rational-responsibility-actional-personal vocabulary by cutting all its ties to a naturalist vocabulary, but you also want your concepts to be useful for the naturalization of language, morality, norms etc.”
Well, I want to distinguish better between the two vocabularies, not to cut the links between them. I still hope that the gap between the two vocabularies could be somehow bridged at the level of some basic concepts. As I have said, my project is not against a naturalist reduction of one vocabulary to the other. What would oppose is just an eliminativist reduction. The point is to be able to formulate clearer and more heuristically fertile problems (and solutions) within the proposed vocabulary, not to prevent a reduction. If the vocabulary proves useful in this sense, then its elimination would be unreasonable.
“I think you are still far from completing such a task.”
I do, too.
“Still, zero-level responsibility cannot simply be the absence of a rejection. No commitment can appear like this. How do you distinguish between ‘being asked for reasons and ignoring the request’ and ‘tacitly accepting the request, while not being able to provide reasons’?”
I suppose you want me to resort to psychological states at this point.
“I do not see other way for you.”
I can look at what counts as an acceptance of the request to provide reasons on the background of different practices.
“You could, but if the acceptance is an act,…”
Ignoring a request is an act. Not ignoring a request does not have to be something in particular. I can look (non-ostentatiously) clueless when you ask me for my reasons for causing A. That could be enough for you to take me as having accepted your request for reasons.
“What if I ask you for your reasons for having caused an earthquake on the other side of the Earth and you look clueless? Should I consider you zero-level responsible for the earthquake?”
“Are you sure your talk of persons does not come with any ontological assumptions? After all, one must have a mind in order to be a person. One must be conscious, able to make decisions and assume some responsibility for her / his actions in order to be a person as well. Do you want to say that all talk about such things does not have any ontological commitments?”
I would say that my use of the term “person” within what I have called the actional-personal vocabulary does not have any psychological import. My use is closer, if you wish, with the legal use, according to which even a company can be treated as a person. Sure, I do not want to say that being a person is reducible to the possibility of appearing as one of the parts in a trial, but I want to suggest that it is not necessary to have a mind in order to be a person, in this sense.
If I attribute actions to X, then X is a person. It can be argued that I cannot attribute an action to X unless X is in principle able to attribute actions to X. However, being able to attribute actions to yourself (or to assume responsibility for your actions) does not need to be reduced to something psychological, as you have already seen. It all seems to come down to taking part in reason requiring and reason providing practices. Companies can be persons if we accept them to take part in such practices. The justification stops at this point. No mind is required in this process of justification. This is the way in which say that I am a person (in this context).
“How could you distinguish between isolated action and joint (or cooperative) action then? When I play a tune together with the other members of a band, my intentions matter. If I do not intend to play the tune together with the others and it is just a coincidence that what I am playing goes along with what the band is playing, I do not cooperate.”
Sure, but why do you need to say that one figures out what one was doing by introspection?
“If you know that you were not paying attention to what another person was doing at t, how could you claim that you were cooperating with that person at t?”
I could cooperate with other members of my soccer team to score a goal even if I wasn’t paying attention all the time to what the others members of the team were doing.
“This does not answer my question. You were part of the team all the time, but you were not performing joint actions all the time. You did so only when you were paying attention to what the others were doing.”
Suppose we move some furniture together, but I do not pay any attention to what you are doing. Couldn’t it be done?
“I do not know what to say. In any case, how could you provide a justification for saying ‘I did A together with Y’ if you do not use any psychological terms?”
Why should need a justification. Couldn’t ‘I did A together with Y’ be similar to a declaration?
“But then you could declare anything you want. Since you do not want to use any natural-causal vocabulary as well, you could, for instance, say that you have played a piano sonata together with the pianist, although you were in the public. (I suppose that you do not want to declare that you took part in an action which you haven’t even witnessed.)”
Sure, I could try to assume zero-level responsibility for having played that sonata, but I wouldn’t be justified to do so.
“And why is that?”
If I did not rehearse it in advance (assuming that rehearsing a sonata is a necessary condition for playing it), then…
“And what if you did? It is still the case that you did not played it together with the pianist at that concert.”
I cannot provide reasons for causing a rendition of the sonata at that concert.
“So what? Suppose you accept to provide them and fail to do so. Your acceptance would still count as assuming zero-level responsibility for having played it.”
But surely the pianist would say that she played the sonata alone.
“It’s her word against yours (according to your view).”
But even if I believed that I did somehow help the pianist to press the piano keys, such a belief with respect to my causal connection to the rendition of the sonata would not justify my claim that I have played it together with the pianist, but only explain why I issue that claim.
“What would your justification be, then?”
Your asking me for a justification with respect to a case in which no reasonable justification could be provided.
“I am asking for your justification in regular cases. However, the justification should be such as to make us sure that your regular cases are different from this irregular case.”
Ok. Then I am going to say that the difference does not reside in a particular feature of my justification for a regular case. It resides on our practices.
“So your claim to have moved the furniture in a joint action with B is justified because it agrees with some regular practice, while your claim to have played the sonata together with some pianist is not justified because it does not agree with our regular practices? Is that your answer?”
No. My answer is that my claim is enough to grant me joint action in the furniture moving case without the need for a justification. The existent practice does not provide a justification, but removes the need for one.
“I am not sure I see any difference here.”
Suppose that I claim to have helped the pianist in the above example. I have rehearsed the sonata and watch the pianist playing, listen to her playing and perhaps anticipate various pressings of the keys. During the play I am focused on what she is doing etc. Suppose, in addition, that the pianist says that she ‘felt as if I was helping her’ and accept this new practice of joint action. Now, what you seem to say is that she would only accept such a joint practice if she was assuming that some of my ‘mental actions’ were causally related to the rendition of that sonata (that could be related, perhaps, to the meaning of the verb ‘to help’).
“Pardon this interruption, but even your distinction between regular duty-responsibility and proxy responsibility needs some causal talk. In the case of proxy responsibility you are responsible for A because, although you did not cause A, you could have affected A.”
I see. Suppose, then, that my psychological makeup is different. I view the things happening around me as if I was in a dream. I see my body moving but I do not have the feeling that “I am causing my body to move”. According to you, this should stop me from assuming any responsibility for pouring some tea into my cup. But it doesn’t. I can continue to say that I have poured tea into my cup. In doing so I am (or can be) guided by the existent practices.
“I know that you think cats do not have the feeling that they cause their body to move (or that they cause things to happen). Suppose you were right and cats were always feeling like people having dreams. Things were happening to them and that was all. Your claim, then, is that cats do not perform actions, but not because of that. The do not perform actions because they lack some practices – namely those of requiring reasons and providing reasons. Is that so?”
Yes, it is. Furthermore, since I have been talking about rudimentary (non-verbal) reasons-requiring and providing practices, I could agree that there are degrees between not being an action performing agent and being an action performing agent. We use a threshold here, of course. If some beings have such rudimentary practices that resemble enough our ones, we could say that they are agents.
“But how could you talk about such practices without assuming that there are minds involved? In talking about practices you employ some ad-hoc social science vocabulary. How could this work without being connected to any psychological vocabulary?”
Let us think to the following case. There are some “entities” which appear to have practices very similar to our practices. The appear to be doing things. Sometimes they appear to ask each other for their reasons to do some things and to provide reasons for doing things. If their practices did resemble our practices well enough, we would perhaps say that these entities are acting agents, responsible for what they do and conclude that they are persons. In this case, most of us would perhaps conclude that these beings also have minds. Even if their inner workings were so different from our inner workings that we couldn’t understand how ‘their minds work’, some people would still claim that they nevertheless have minds.
What I am trying to say is that we are not compelled to conclude that those persons have minds. If we had the strongest reasons to believe that those beings are robots, for instance, we could say that those beings, who are persons (since they perform actions in the ‘space of reason’) do not have minds. Of course, it might be the case that we ourselves are such beings. If you are an eliminative materialist (as I tend to be), you will agree that you do not have a mind (subjective, qualia-like experiences could be perhaps described as empirical actions performed on your own brain in this case – this is a different problem, anyway), but that would not necessary mean that you are not a person.
“You have avoided my question.”
Not at all. There is a graded series of cases which we could describe by speaking of events, practices, social practices and finally ‘interpersonal practices’. Even if we tend to attribute ‘interpersonal practices’ (like that of asking for reasons and providing reasons) to humans only, we could attribute some social practices to animals and we could also attribute practices to beings of which we do not believe that have minds (like robots). We associate some practices with humans only and say have said for centuries that only humans have minds, but that is a historical accident. It imposes no constraints on the way in which we use our concepts. Our ancestors, after all, could have been speaking of the sea and the wind being involved in some practices (playing some games). We believe that in talking like this they were attributing a mind (psychological states) to the sea and the wind, but what if they were not? In any case, nothing prevents us from adopting this talk with respect to other living beings. Our practices must have evolved from theirs, after all.
“Right, but animal practices are more rudimentary than human practices because animals have more rudimentary minds.”
Why not simply say that evolution produces complexity? You do not need minds in order to explain why our practices are more sophisticated than their practices. On the other hand, if one gives up the “mind myth”, several philosophical conundrums are avoided: body-mind, mental causation, other minds, solipsism – to name just a few.
“Sure, but we are still left with this double talk: ‘I am an animal body causing changes in its environment’ / ‘I am a rational and responsible person performing actions’. How does this solve anything? The second kind of talk can be used to reject solipsism only if you ‘inject an ontology’ into it. There cannot be the case that you alone exist because you perform actions and those actions are events which need to effect changes in an environment (since otherwise you could not have purposes, that is ‘yet uneffected changes’. But if a person is not ‘something existent’ and an action is not ‘something existent’ and there are no causal relations (between what a person does and what a person causes to happen, for instance), solipsism could as well be true.”
Perhaps my attempt to use the actional-personal vocabulary to argue against solipsism was wrong. Perhaps we do not need to argue against solipsism if we actually get involved in the regular practices in which we use the naturalist vocabulary.
Most of the old problems were due to an ontological clash between the naturalist vocabulary and the mentalist talk. But if the personal-actional talk has no ontological commitment, there cannot be any clash between this vocabulary and the mentalist one.
“This is weird. You seem to claim that ‘X exists’ does not follow from ‘X is a person’. Now, let’s take an example with God. One could say that God is a person and refrain from saying anything about the existence of God, but that could only mean that the speaker says ‘If God existed, God would be a person’ (it’s the same with ‘God has necessary existence’). But now you say that one could say ‘God is a person’ in a different way and still deny that ‘God exists’. There is no middle ground between ontologically committed talk and conceptual talk. Or is it?”
I have to think about this. Let me start with a different case: “You could marry Adele only if Adele was a person and she is a person. So you could marry her.”. Suppose I say this without knowing whether Adele is still alive. When I say that Adele is a person I do not want to be ontologically committed to Adele’s existence. I do not express some conceptual truth (as in ”Adele’ can only be the name of a person’). I am perhaps saying that Adele may take part in our interpersonal practices. She may provide her reasons for performing actions, for instance. Now, I could say the same thing about Sherlock Holmes. This could be also have the form of a counterfactual: “If Sherlock Holmes existed, he could take part in our interpersonal practices (we wouldn’t forbid him from doing so).”
The difference, perhaps, could be that while the consequent of the first counterfactual looks like a statement (a constatative), the consequent of the second counterfactual is of a different kind.
“Still, you could at best talk about ‘practice recommendations’ being associated with a concept. How is this not a conceptual matter?”
My analysis could be considered to be dealing with conceptual matters, of course. But now think about it. ‘God is a person’ comes down to issuing some ‘practice recommendations’ (to use your phrase) with respect to God, whether God exists or not. Nevertheless, I can actually engage in such practices towards God without being sure that God exists. So the practice recommendations were not empty.
“I suppose the recommendations have the form of ‘you ought to…’ (you have talked only about ‘you may…’). Leaving the weirdness of engaging into a practice with a nonexistent person aside, such recommendations should have a conditional form, when spelled out. I suppose their antecedent will say something about what the other person did or some occurring circumstances in which the other person is involved (e.g.: ‘If a person X asks you for your reasons to do A, which was addressed to X, then you ought to either provide your reasons, or admit that A was not reasonable.’) So how can such practices be conceivable?”
Even an atheist could continue to honor her parents after their death. I assume that ‘being God’ means ‘being responsible for all that is going on around’. If I am happy with what is going on around me I feel like thanking the person responsible, so I thank God for what is going on.
“And you can do that knowing that God could as well not exist?”
Well, I think it is reasonable to believe that God does exist. I would not say that God exists as a physical body in nature, but God could have transcendent existence (nature is like a computer game and God is not inside the game, although God could enter into the game as one of the players). This is, however, besides the point. Thanking the person responsible for something beneficial to you might not be a moral obligation, but it is polite to do so. So if one wanted to be polite, one could do it without making sure that the person in case does actually exist.
“See, even you accept to talk about a person existing. This shows that talk about persons does not exclude any ontological commitment.”
I did not say that the personal-actional vocabulary is such that it makes ontological commitment meaningless. My claim was only that such a vocabulary could be developed (or analyzed) without looking at any ontological commitments.
“But this only means that we could combine the two vocabularies and get a clash between their different ontologies, doesn’t it?”
Oh, well, this could happen, perhaps, but then we’ll always have a simple solution at hand in such cases.
“So you are allowed to talk about persons causing events, but when I ask you how could a person, which does not seem to be the same thing with a body, cause the occurrence of a physical event, you are going to reply that a person does not have to be something and so on? How could this satisfy anyone’s intellectual needs?, I wonder.”
You got me wrong. I could rephrase my utterance by saying that it was a body which caused the occurrence of a physical event and a person which was responsible for the occurrence of that event and while the first conjunct in this reformulation states a fact, the second makes some ‘practice recommendations’.
Perhaps we talk like this (‘Person A did something harmful to me’) because we usually need to say both what happened and to propose some actions. I don’t know. Sometimes, the fact stating talk seems incomplete. ‘There is a cup on my table.’ – So what? What should I make of this? Why am I saying this? If I am saying this to you, perhaps I want to suggest you to do something (take responsibility for putting it here, tell me why, take it back etc.) or to let you know that you are allowed to do something (use the cup, ask me why I put it there etc.). I am telling you this fact because something should (or could) be done about it.
The double talk, in itself, might not be double at all. It could be that we can only make explicit what we say by using such a double talk in our metalanguage, and we do this only to avoid some conceptual problems. The question is: ‘Do these problems require us to improve our concepts or only to clarify them?’. It might be that we do not need a unification of the ‘metalanguage double talk’…
“Are you sure you want to talk about metalanguage here? In any case, how are you going to say how does language relate to reality if you stop at this ‘double talk’ hypothesis? You wanted to talk about actions (empirical, communicative etc.) precisely because talk about sentences ‘floating in logical space’ seemed unsatisfactory to you. Now, however, your ‘actions’ seem to float in ‘the space of reason’ in rather the same fashion. What was gained?”
Let’s take an example. Suppose that I tell you that there is a cup on my desk. This enables you to ask me a lot of questions: What color is it?, Is there anything in it?, Who put it there? etc. Also, I would not be able to tell you the truth unless the action of putting a cup on the desk was performed by somebody. Or at least, my saying that there is a cup on the desk is made possible by watching the cup on the desk (or some other empirical actions). If you try to take the cup on my desk (or just remove it) and fail, you are going reject my utterance (that is, you will do a communicative action opposing mine). All this network of actions is what gives the content of my sentence. By including ‘non-communicative actions performed in an environment’ in this network, the link between language (or communication) and ‘reality’ is made possible. Now, you can contrast this to the traditional representationalist view: the word cup (of which the word uttered by me is just a token stands, via some mental representation or abstract conceptual content for the cup on the desk, the phrase ‘my desk’ refers to the desk in case and ‘on’ stands in some obscure way for the relation between the cup and the desk. The entire sentence describes the existent situation via a representation (which can have a non-apparent logical form) and some abstract functional relations of reference.
“But your actions have nothing to do with the cup, since the cup is a physical object and actions, as you say, do not exist. Also, the cup could have been glued to the desk, so the failure to remove it from the desk would not mean that there is no cup on the desk.”
I take your second point. As for the first one, I was trying to say that my talk of actions, persons, intentions, responsibility and the like does not have any ontological commitment.
“But it surely does. You have been talking of events and persons causing events in order to say what is it for someone to perform an action.”
Ok, then perhaps there is something wrong in my claim that actional talk has no ontological import. I wanted, however, to leave open the possibility to talk about a person performing an action in the absence of an occurring event. We might want to say, for instance, that if Smith has changed his views on a subject, this is something which Smith did. Smith can be held responsible for having changed his views even if cannot say what was the event called ‘the change of Smith’s views on the subject of…”. We do not have to talk about ‘mental events’ in order to say that Smith is responsible for having changed his views. On the other hand, Smith cannot be responsible for a neurophysiological event having occurred in his brain.
“So you think the solution is to cut the ontological commitment altogether?”
This is what I have thought, indeed. Mental actions do not have to be mental events.
“But why cannot they be neurophysiological events? The brain is a tool. You do not need to understand the functioning of a tool in order to use it. Why couldn’t we say that Smith has changed his views by using his brain?”
Well, you do not need to understand the functioning of a microwave oven in order to use it, but you must be able to control the functioning of the oven in order to use it. It is awkward to say that Smith is using his brain to change his views, because Smith cannot control the functioning of his brain.
“Why not? We seem to be able to control the functioning of our brains quite well when we use it to talk. When you talk, you can stop talking at any point. It is not as if talking is happening to you. Talking is something which you do. You wouldn’t be able to do it without a brain, so you are using your brain to do it.”
I wouldn’t be able to breathe if Earth did not have an atmosphere. This doesn’t mean, however, that I am using the Earth’s atmosphere (as a tool) in order to breathe.
To make a long story short, using a brain is too dis-analogous to using a tool (a computer, for instance). There is a different sense of “using” something in which we can be said to use our brains. When you use a tool, what you do with the tool can still be described as an action (inputting something in the computer by using the keyboard, pressing a key on the keyboard, reading the computer screen etc.). If you use a hammer, you can say that you are holding the hammer, thrusting it forward towards the nail etc. When you use your lungs to inhale some air, there is no action you do with your lungs. This is a different meaning of “using something”. We can also use a supply of raw materials to build something – this is yet another meaning of “using something”. The list is perhaps open (the object you are acting upon can also be said to be used). If we are interested only in “using the brain as a tool”, then it can safely be said that we do not use our brains. Similarly, muscle contractions are necessary for a lot of actions, but except for the cases in which we are responsible for contracting our muscles and accept to give reasons for doing so, we do not say that we use our muscles (as tools) in order to do other things (run, walk, grasp objects, lift them etc.).
Also, by using a tool someone does something else. I thrust the hammer forward in order to hammer a nail. ‘Nail hammering’ is a different event from ‘forward hammer thrusting’. Even if we could describe ‘changing something in one’s own brain’ as an action, Smith would be doing that in order to change his views. One could replay that ‘Smith has changed his views’ is just another description of the same event, but then Smith couldn’t have used his brain as a tool. A tool is used on something.
“Maybe Smith’s brain was the object acted upon, not a tool.”
A neurosurgeon can act on somebody’s brain. What Smith does is in no way similar to such a case.
I would rather say that there are no mental actions than accept that these are actions we perform on our brains.
“What about noticing that you feel a particular pain by introspection? Isn’t this an empirical action performed (internally) on you brain, after all?”
I think Wittgenstein has said a lot about such cases. Let’s take an example of a special qualia – the experience of cutting something with a ceramic knife. One cannot have it if one did never cut something with a ceramic knife. I can say: ‘Cutting a fruit with a ceramic knife does not feel like cutting, in a way. It feels like making a simple gesture and producing a slice of fruit.’
To have this qualia is to perform the action. If you did not do it, I can indicate another action – cutting a fruit buy pointing you finger at the place where you want a cut to be made. By performing this other action you can perhaps have the same qualia. It is assumed, of course, that you body is functioning normally. Why not say that the ‘seeing red’ qualia is the experience of performing the action of looking at a red object when your body (including your brain) is functioning normally. Major differences between bodies (between my body and the body of a bat, for instance) can prevent us from saying that we have similar experiences when we perform similar actions, but most of the time this is not a problem. A color blind person cannot look at red things in the same way as I do. She must use her reason (in a different sense of “use”) in order to identify a red thing to look at, I can “use” my sensory input only (it can be said that this is the ‘raw material’ out of which I produce a red object by looking, but this is perhaps not the best way to describe what I am doing). However, a color blind person can look at a round object in the same way in which I look at a red object. If all the round objects were red, the color blind person would look at red objects in a way quite similar to mine.
“For such a person ‘red’ would still mean round.”
Looking at red things would be done by the color blind person by taking shape cues from the sensory input into account. That is true. However, having the sensory input of a color or a shape are not empirical actions. You can, perhaps, be said to look at the shape of an object or at the color of an object. A raw sensory input, however, is what we sometimes call an unconscious sensation. Sensations produced by peripheral vision, which are used to adjust walking on an irregular terrain while talking to somebody are such a raw sensory input. Somebody could determine that shape cues matter more the color cues in the adjustment of our walking in such a case, but if the color were to matter more, our experience of waking on an irregular terrain while not paying attention to what we are doing would not change.
What I want to say is that we can consider a qualia to be the experience of performing a certain empirical action. If the same empirical action is performed in the same way, then the experience of performing it can be said to be the same.
“Isn’t the experience subjective, though?”
Well, I do not want to use it to refer to some private subjective mental state or event. Think of the question “Do you have some experience with wielding?” The question is not about having been in some special subjective state. If you say “I did wield something one night, when i was drunk”, I might conclude that you do not have experience at wielding, after all. Now, suppose that I was asking you: “Do you have any experience with smelling garlic?”. If you say that you did smell some garlic once, when you had a cold, I will deny that you have any experience with smelling garlic, not because you did not have the particular subjective experience of “feeling the smell of garlic” (since you were unable to feel the smell due to the cold), but because I think that you cannot gain any experience with smelling garlic when you have a cold (in the same way in which you cannot gain experience with wielding when you are drunk).
“What if the garlic I did smell was bad and did not produce the regular garlic smell?”
The you do not have experience with smelling regular garlic.
“What if we talk about how it feels like to smell garlic for a cat?”
Cats can perform empirical actions too. It might be that they only have rudimentary concepts (since any action has a conceptual structure), but think about a non-linguistic dream in which you smell garlic. Since it’s a regular dream, it will not feel like you chose to do something (I suppose cats do not chose to do whatever they do). Also, you will not use language and perhaps the conceptual structure of what you experience doing in your dream will be quite rudimentary. It could still be said, perhaps, that you have gained the experience of smelling garlic as a cat.
“But how can I be sure about that?”
I suppose that if one compared what happened in the brain of a cat when the cat smells garlic with what happened in the brain of human who dreams of smelling garlic (perhaps because some garlic is present in the room in which she is sleeping) and what happened was similar, we could make sure about that. In any case, if we had means to temporarily make the functioning of a human brain to massively resemble the functioning of a cat brain, then the human subject for this experiment could experience smelling garlic as a cat. There is no need to talk about subjective experiences for this.
However, we seem to have strayed from our initial discussion. Actions can affect objects. Thus, they have to be something. I was ready to agree to this point.
“Right. So how do your so-called empirical actions affect the objects acted upon? You do not change a chair by looking at it.”
The object acted upon does not have to change. I can give an object to you without modifying the object in any way. I do not even need to move it in order to give it to you. I could, for instance leave it to you in my will and it will become yours after my death. Something was affected – my will, some property rights, if you wish – but the object acted upon did not change. When I look at a chair, the focus of my attention changes.
“And doesn’t this change amount to a mental event?”
Not necessarily. It could amount to a neurophysiological event. Not all events must be actions. Not all the physiological events underlying my grasping the cup on my desk are my actions (in short, basic actions are the most simple events underlying one of my actions which I am willing to provide reasons for having caused – for instance, I might be willing to provide you with my reasons for some preparatory movements or a part of my hand movement in throwing a ball, but not for the movement of my hand over a tenth of a second of that throw; I could tell you why I started to throw the ball from a low position, but not why I didn’t release it one tenth of a second earlier).
“So you want to say that neurophysiological events are not actions, but they can underlie some actions.”
I do not think that it is in principle impossible for someone to learn to produce a particular change in one’s brain (perhaps in the same way in which one learns to move an artificial limb). There are no limits to what actions we can perform. It is in principle possible to move or bend objects without moving your body or contracting your muscles (and I am not talking about psychokinesis here). We are just not able to do it yet. One could do it, perhaps, with some training, by using a brain interface wirelessly connected to a robot. We can already use low level instructions in an assembly language to put particular bits in a computer’s particular memory location. I am inclined to say that things are the other way arround: an event is an action which we still cannot assume responsibility for. Sellars arguments about ‘looks like’ talk depending on ‘is’ talk could be used here as well. We know how to witness ‘natural’ events because we know how to act and then we learn to withdraw some commitments from some possible actions (the same could be said about beliefs: expressing beliefs or opining that … is something one can do only if one knows to make statements and to withdraw some commitment from a statement).
“That’s a lot of speculation.”
Perhaps. Also, I am using ‘assuming a responsibility’ and ‘making a commitment’ interchangeably here. For instance, if I promise to do A to you it could be said that I have undertaken a commitment to do A and also that I assume responsibility for having promised to you to do A. If I order you to do A, I am not undertaking a commitment to do anything, but I assume the responsibility of having ordered you to do A and perhaps I also assume some kind of proxy-responsibility for A.
“This isn’t a conversation anymore. You just throw in one idea after another.”
Perhaps the topic of my book could be expressed by the following question: How does language mesh with our life?
The suggestion is that the two are interlocked or interconnected. The problem is that they are so both in the ‘space of reason’ and in the ‘space of nature’. This is not just the old dispute between empiricism and rationalism (or idealism) in new clothes. We live our lives both as part of nature and as rational persons. Our lives are composed of our actions (some of which are performed by using a language). Thus, our actions must enter both into causal relations (with other actions and also with natural events) and into rational entailment relations. Two questions emerge:
1) How could we better conceptualize the two interlockings?
2) How could we unify our conceptual frameworks for both interlockings?
Now, this might be too big a project for just one book. I could start with my conceptual proposal for the interlocking between language-use actions and other actions (and perhaps natural events) in the space of reason.
Here I have several ideas which could be used (and perhaps developed):
– my concept of basic reference (tagging or predication performed directly on objects acted upon by an addressee etc.; the relation between implicitly communicative actions and communicative actions)
– my concept of empirical action (looking at objects, touching them, watching events, measuring things etc.)
– my concept of opposing actions (which indicates a logical link between communicative actions and non-communicative actions)
– my idea of rule embedding artifacts
– an analysis of the concept of ‘rational entailment’ as ‘transmission of responsibility’ (sketched above)
– an analysis of the concept of ‘performing an action’ (sketched above)
– other developments; see for instance the problem: How can you have action entailment if there are no conditional actions? (my paper on logical form partially answers this); another problem: how could you have some entailment resembling inductive generalization with non-communicative actions only? another problem: how can one distinguish between logical implication (or some rational entailment similar to logical implication) and implicature if entailment is conceived as ‘transmission of responsibility’? etc.
– applications to cases (treatment of some existing problems from literature)
Perhaps it would be better to focus on these matters for now.
An action requires:
– an agent (a person; able to assume (zero-level) responsibility for actions / to undertake commitments (duty-responsibilities) for performing actions)
– an addressee (or beneficiary; this could be any being capable of feeling good or bad; a pet, etc.)
– an overseer (a person entitled to ask the agent to provide her reasons for performing the action)
– a purpose (an outcome; thus, actions seem to require an environment)
– something acted upon (maybe)
Perhaps a version of the private language argument could be used (since the concept of ‘reason for action’ is involved) to say that the overseer must be (in principle) distinct from the agent – so the agent must be able to recognize someone else as a person
– being able to recognize X as a person (potential agent) vs. treating X as a person;
– accepting the request to provide reasons (as justified) vs. accepting to provide reasons to a particular person;
“Isn’t a sentence formed naturally, by erosion of a stone wall, still a sentence? Shouldn’t you care about sentences and not actions?”
If it is not addressed to me, why should I care about it. And if it isn’t addressed to anybody at all, why should anyone care about it? You can call it a sentence, if you wish, but I do not think language works like that. I do not think that something can be said (or written) without an addressee. In addition, only an action can have an addressee. Natural events do not have any addressee (unless you consider one a sign from God etc.).
Autobiographical memory as reenactment of empirical (and non-empirical) actions. […]
“But you must know which actions to reenact, so you must represent them. But then, those representations are enough to remember everything. So they are your memories.”
Actually, what I am saying is that autobiographical is more like motor memory (think of remembering how to play a tune on a musical instrument).
Every action has an object acted upon. When I simply move my hand away from the lit cigarette I have accidentally touched, the object acted upon is my hand. When I lift a weight, the object acted upon is the weight (now the hand is ‘transparent’; using Heidegger’s distinction between Vorhandenheit and Zuhandenheit, one could say that the hand is ready-to-hand). When I hammer a nail, the nail is the object acted upon (now the hammer is ready-to-hand, together with the hand).
Now, any particular action has a particular object acted upon. The action type, however – ‘hammering a nail’, for instance – does not have a particular object acted upon. Thus, the action type can be called a universal action. We do not perform action types, but particular actions, of course, but in speech we seem to be able to utter universal statements (“All humans are mammals”). Think of the action of tagging a particular human as mammal. That is an action performed on a particular object. Now, the action type is ‘tagging a human as mammal’. It is interesting that we talk as if we were performing the action type in this case.
By tagging X as ‘food’ you “insert” it into a range of possible actions – eating X, feeding X to D (an addressee), cooking X (if it is raw food), storing X in a refrigerator, chewing X, classifying it further as some kind of food, etc.; some actions seem conceptually linked with the tag used, others do not;
It’s the same with all concepts (we could say that).
“Right, but if you started by thinking of sufficient conditions, you would have said that making tea is the rational consequence of boiling the water and putting the teabag in it and not the other way around.”
That is because the two (making tea, on one hand, and boiling water and putting a tea in the boiling water, on the other) are equivalent. Making tea is not an extra action. It consists in those two other actions. By contrast, unlocking a door is not sufficient for opening it.
Thus, opening a door does not follow (by reason) from unlocking it. It is true that one unlocks a door in order to open it, but no rational relation can be established between those two actions.
On the other hand, since one can open a locked door only if the door is unlocked first, we could say that the action of unlocking the door is a rational consequence of the action of opening a locked door (this means “opening a door that was locked”).
“What if the door is unlocked by a mechanism with a sensor activated by my presence in front of the door? Do you want to say that one unlocks the door by standing in front of it?”
Not necessarily. Here we have a choice. We could say, for a while that by standing in a certain spot in front of the door in order to unlock it is an action, but after a while we could only talk about opening and entering through such doors as actions. We might say that the doors allow only certain people to open them. The rules with respect to who could open a door would be conceived as a part of our environment. Perhaps we will not say that the doors are locked and need to be unlocked anymore.
“But in order to open a door you might need to be, let’s say, the president. Wouldn’t being the president be a necessary condition for opening the door, even if ‘being the president’ is not an action?”
Sure, but talk of “necessary conditions” in such cases stands for causal talk. I could jump 10 meters high if Earth’s gravity was lower. One could express this by saying that Earth’s gravity being lower than x is a necessary condition for someone weighting y kilos jumping 10 meters high. There is a causal connection between the two, such that the second (conceived as an event and not as an action) is not possibly unless the first (which you could call a physical state) obtains. Some events can only occur under some physical conditions. That is all.
“Anyway, you want to conceive those rational consequences in terms of responsibility, desho? By unlocking a door, then, one seems to be responsible for the door being opened at a later time.”
I am not necessarily zero-level responsible for opening the door, since somebody else could do it.
“But then, by performing the final action in a series you are not zero-level responsible for other actions in that series (even if they were necessary for the performance of that final action), since those actions could have been performed by somebody else.”
Please ignore my previous reply. I should have talked of duty-responsibility. By unlocking the door I do not become duty-responsible for opening it. By uttering a sentence I become duty responsible for accepting the logical consequences of that sentence. Somebody else could accept one of those sentences on my behalf, of course. In the same way, if I am preparing a tea, somebody could boil the water for me. This does not mean that I was not duty-responsible for boiling the water.
To sum up, it is easier to use a new artifact if it resembles an existent one (examples: the electric toothbrush vs. something completely different to clean your teeth, the vacuum cleaner etc.). Not only does the new artifact have to resemble some existent one, but it has to fit with other existing artifacts. It is the same with concepts. My personal-actional vocabulary is a new set of tools, in this respect. I am not trying to say what an action is, but to propose a new concept of action. It’s main use is within a unified theoretical account of rationality which, in it’s turn, should provide a rationalist theory of meaning and reference, should underlie (most of) the existent logical theories (illocutionary logic included) and should not antagonize naturalism, but avoid eliminativism. However, my concept of action (and the related concepts of responsibility, “being a person”, etc.) should resemble existent concepts and fit with (at least some) common and theoretical linguistic practices. This is what makes the whole project so difficult. The point is that I see no other way in which our conceptual means could be improved. If the project is doomed, then I am doomed as well.
“Well, I wouldn’t say that the project is doomed, but I cannot see that it has good chances of success either. Since my intellectual abilities cannot surpass your own abilities, I cannot suggest another direction of research to you, of course. After all, I am just an imaginary opponent.”