22.09.2014

The project of my paper (1):

Empiricism: All knowledge about nature is formulated in sentences which must be justified by sentences describing our experiences.

Wittgenstein's PLA: If by "sentences describing our experiences" we understand sentences in which we describe / or refer to some inner experiences, inaccesible to other language speakers, then such sentences cannot be a part of our language. So they cannot be the ground for all our knowledge about nature.

Argument (Sentences describing inner experiences cannot be a part of our language):
Suppose I was to use a sign 'S' to refer to a particular sensation S. There is no way in which I could learn this sign by ostension from another person. So I must be able to introduce it myself with an ostensive definition. 

Two steps:
(1) Introduction of the sign.
(2) Using the sign correctly.

In order to say that I am using 'S' correctly to refer to S I must have some criteria for correctness, assuring me that the sensation to which I refer now by using 'S' is the same kind of sensation as S (as I remember it). It must be in principle possible for me to be wrong (remembering S incorrectly is one way in which I could be wrong, but not the only one). But I cannot have such criteria if it is in principle impossible for someone else to correct me if I am wrong. (this does not have something to do directly with the rule-following considerations - those apply to a discussion about the meaning of the words; here the relation of reference is at stake - see also Russell's authentic proper names, which do not refer via a description; instead of 'S' we could as well talk about 'this sensation' and 'that sensations' and criteria for identifying and distinguishing sensations)

[Sure we say that we distinguish visual sensations from hearing sensations. We also believe that we could distinguish such sensations even if we knew nothing about our sensory apparatus - they would still "feel different". I do not want to enter into this right now. We could think about distinguishing the sensation of hearing a sound (A) from the sensation of hearing a different sound (B).]

[Having some sensory input is a private experience. Listening to a sound made by a piano is a public activity.]

Since there is no way in which I could use the sign correctly or incorrectly, I cannot introduce it.

Main discussion: why are there no criteria for distinguishing correct from incorrect uses of 'S'? And why do we have such criteria for the signs we use to talk about publicly accessible objects?

Sellars:

Suppose one would describe his/her experiences not as some private inner objects (or events) but as public ones. I could say 'There is a tree in front of me' instead of 'It appears to me that there is a tree in front of me'. If Wittgenstein has a problem with sentences which seem to describe inner mental objects we can drop them, but keep in mind that we are still describing experiences. The real problem is that no such experiences can ground our knowledge, since our knowledge has a conceptual structure (grounding also depends on logical relations, which only hold between sentences, which must have conceptual structure), while the said experience do not have a conceptual structure.

McDowell: If they must have conceptual structure, then our experiences have a conceptual structure. This is an epistemological, not a psychological point.

A whole discussion after that.

Back to Wittgenstein. He seems to discuss about justifying our knowledge and replying to skeptical objections in On Certainty. One possible reply is suggested there: the starting point of all our justifications (with respect to the knowledge of our environment) are our actions (we act in an environment). 

My point: if we conceive the way in which we "have experiences" as a way of acting in our environment, we might escape skepticism with respect to empirical knowledge. To know things from experience is to state sentences (also an action) within a series of cognitive actions which include our interacting with our environment. Our actions can figure in the space of reasons (Sellars), since they have a conceptual structure (Anscombe) and enter into logical relations (my point). This view should also enable us to talk about epistemic virtues (knowledge producing actions are those performed by agents who have developed epistemic virtues). A way to bridge the gap between the "space of reasons" talk and the naturalist talk should be devised, but this can be done some other time. It is also interesting to see that the reply suggested in On Certainty brings Wittgenstein quite close to the phenomenological tradition.

*

The project of my paper (2):

- State Wittgenstein's PLA in its best conceivable form (without entering into exegetical debates). Make it as convincing as you can (even if this means to disregard the resolute-Pyrrhonian tradition; Reason: I am interested in a philosophical debate, not in wittgensteineanism).

- Clearly point out the weight of this argument against the logical empiricist tradition (three main attacks of that tradition: Quine, Wittgenstein, Sellars). 

- Talk about how Sellars understood the argument and why he thought his argument was better and deeper than PLA.

- Clear missunderstandings, point out to the suggested sollution in On Certainty.

- Show why the suggested solution should be developed and point out why it has good chances to save our empiricist intuitions in a coherent conceptual framework.

*

I climb a tree only if I (non-psychologically) know that that tree exists. 
If I climb a tree, then I know (not as a psychological state) that that tree exists.

I watch a tree (I perform this action, under this description) only if I know that that tree exists. 
If I watch a tree, then I know that that tree exists. 

I watch a bird. Someone asks me: "What were you doing?". I answer: "I was watching that bird". She says: "What bird? There are no birds here". I reply: "Then parhaps I was watching something which looked like a bird". 

I redescribe my "perceptive action". However, I do not give up the assumption that I am in an environment and act against it.

So I  do not need to justify the existence of an external world. I act (even perceptively) in an environment. This is not external. My actions, when performed in the appropriate epistemic context, prodece knowledge about my environment. That is, they are followed by speech acts of the type "That is a tree". They are also reasons for asserting that "That is a tree". 

23.09

I am told that some object has fresh paint on it. I touch it to check this. Did I know that the object existed when I touched it. I am inclined to say yes. Did I know it when I reached to it? Perhaps, but here I am more inclined to say that I assumed it. So what is the difference between knowing that there is a tree there and knowing that the tree has some fresh paint on it? 

Perhaps I could say: "Well, I did not reach to touch the tree, but the paint. The existence of the tree was not at stake, the wetness of the paint was."

So perhaps touching the paint justifies my belief that the paint is wet in a different way from the one in which it" justifies" the belief that the tree is there.

The paint on the tree is wet on touch only if there's a tree there. 
If the paint on the tree is wet on touch, then the paint on the tree is wet. 

29.09

Several "private language arguments":

243: Private language: a language in which I speak about something accessible only to me - immediate private sensations.

-> in principle inaccesible to anyone else: well, we can imagine two minds sharing the same body, taking turns in controlling it (paralel processing in the brain etc.); while one mind controlls the body, the other senses what the sense organs of the body provide, reflects on this and that, sleeps etc. This can be conceived without a serious contradiction; now, if the two minds direct their attention to the same sensory input, the two persons have the same sensation (we can also assume that th e brain functions responsible for sensory input processing are shared, up to a point etc.) I do not think that on of the two persons would doubt that the other has the same sensation; 

We could consider this scenario a natural extension of the case in which we have sensations. If so, then immediate private sensations are not 'in principle inaccessible to someone else'; they just are not accessible as they are. 

Does the PLA fail if we think about this? - Could the two persons in my scenario (inspired by The Host) develop a so-called private language, which only the two of them would understand?

246: It makes no sense to say that I know that I feel X. (see OC also)

253: Sensations have no criteria of identity. (or at least, belonging to X is not a criteria of identity for a sensation - perhaps it is similar to spatio-temporal location in the case of physical objects - being "located" in X's mind, the sensation cannot be located in the mind of another person - that is what helps us hold onto the impressions that sensations, too, can be particulars (individuals);

See abstract objects - they are not located in a mind and also they do not have spatio-temporal properties, so it is weird to think of them as "particulars". (But we could also think of them as 'absolute particulars' - they can only be particulars, concepts do not form similarity classes - we can have a sensation 'of the type T', but not a concept 'of the type T' - we can only have particular concepts [...])

256: For a private language to be "really private", the name of a sensation should not substitute the natural expression of that sensation, but refer to it directly.

257: To name a sensation one must have learned the game of naming (and she must also be able to extend that game for the case of sensations)

258: To introduce a name for a sensation one has to do it by ostension. But we lack the criteria to call a private ostension "an ostension".

I call O an ostension because I say that O "brings it about that I remember the connexion right in the future."

"But in the present case I have no criterion of correctness" (var1: "for what I do now" - 1a - naming sensation S by 'S' / 1b - naming O an ostension;  var2: "I have no criterion of correctness for the future." - 2a "I do not know what would count as a correct naming of S by 'S' in the future - the future application of 'S' is not determined, fully determined etc.  - see RFC applied to the case of reference, critique of ostensive definitions, of the idea that all applications of a concept are determined etc./ 2b " by focusing on THIS sensation and calling it 'S' I did not produce any criteria for how to use 'S' in the future)

Also, if it only "brings it about that", it is not necessarily something related to language (we only have a causal link here).

259: In order to have criteria we must have rules. Imagined rules are not proper rules. Assuming that a definition by ostension can introduce rules, a definition by private ostension can only introduce imagined rules.

What is the difference between rules and imagined rules? [...]

260: "Perhaps you believe that you believe it." - if the basis for calling a sensation 'S' (later) is your belief, conceived as a private mental state, then we have only moved from naming a sensation 'S' to describing a belief as 'the belief that this is sensation S again'; (see Sellars about unconceptualized experiences being epistemologically irrelevant)

Thus introduced, 'S' has no function. (I want to keep a diary about the recurrence of a certain sensation - that was the intended function of 'S'; but 'S' cannot perform this function by naming the sensation (no matter how it was produced).

261: the kind S belongs to (sensation) is spoken of in our common language and not in a private language -> if we want to continue we are reduced to 'THIS' - as a name for a private mental occurence for which we do not specify a type;

262-4: Supposing that the ostension is successful and that it produces the connection between a name and a mental occurence (this could be disputed see the weird baptism case - each member of the community imitates the priest and says 'I call this child A, B, C etc.', but the names are not used afterwards -, the "temporary name" and other cases), it can still be said that with this we do not know how to use 'S' in all the future  situations (we have a reference, but not necessarily all the rules for the use of 'S')

265: "If the mental image of the time-table could not itself be tested for correctness" - 

"This is S again." - How do you know? - "I remember the moment when I had the sensation which I have called 'S' and this now is similar to that sensation."

And how do you know that you have called that sensation 'S' correctly? (how are we to test for correctness that moment which you remember now, even if we assume that you remember it right) - "Well, I did, because I have managed to impress the connexion between 'S' and that sensation on myself." - But you did that only if know you are able to use 'S' correctly.

Contrast this with the case of quus: "I remember to have used plus and not quus in the past." - How do you know? - There is no memory (or other mental state) which could help you establish that you intended to use plus and not quus -> it's not that we appeal to a memory but it cannot be tested; in that case there is nothing we can appeal to; If I say: "I remember having written '5' as the result of the sum '2 + 3' the opponent does not doubt that. The problem is that we remember nothing about the big numbers which we have never added before.

Bad ostension - one which does not establish the connexion between the name 'S' and instances of the same sensation - S.

The several ways in which the ostension can fail in the case of TOFF (see the Blue Book) can be corrected, after all. 

268: One cannot give a private definition of a word to oneself in the same way in which one cannot give money to oneself. -> this suggests that the problem is not with sensations, but with private definitions (no matter for what things, even if they have identity criteria);

271: A person feels what she calls 'pain' in the same conditions as we do, but she does not feel the same thing all the time. -> If we are inclined to say that this person uses the word 'pain' in the same way as us, then...

275: When we speak about colour a.s.o. we usually speak about "real objects", not about sensations

276: Our talk about the "sensation of red" is based on our talk about red things. (Sellars says something similar)

[Why does Wittgenstein talk about pains first and colours afterwards?]

280: A picture representing a theatre scene and a mental image of that theatre scene in the mind of its author.

283: "[...] Only of what behaves like a human being can one say that it _has_ pains."

-> Animals are in pain. They do not have _their_ pains - ownership.

288: If he no said, for example: "Oh, I know what 'pain' means; what I don't know is whether _this_, that I have now, is pain."...

This is too weird, Wittgenstein says. 

289: "When I say 'I am in pain' I am at any rate justified _before myself_." - What does that mean? Does it mean: "If someone else could know what I am calling 'pain', he would admit that I was using the word correctly"?
To use a word without a justification does not mean to use it without right.

-> So there is no justification. But still, I have the right to say that I am in pain.

291: What do we call "descriptions". - instruments for particular uses (not only pictures are descriptions; in fact, they are not, since they are idle);

292: you would have to apply the rule in the particular case without guidance

293: If we construe the grammar of the expression of sensation on the model of 'object and designation' the object drops out of consideration as irrelevant.

301: (after speaking about the image of pain in our language-game): An image is not a picture, but a picture can correspond to it.

30.09

272: The essential thing about private experience ... The assumption would thus be possible - though unverifiable - that one section of mankind had one sensation of red and another section another.

Right, but the usual logical empiricist (positivist) answer is that since this assumption is unverifiable, it does not have cognitive sense (see Schlick). It follows from this that knowledge has nothing to do with qualia. No qualia can be communicated. We cannot know a thing about how the world "feels like", only about some mathematical structure of it's regularities (as in natural sciences etc.).

In a sense, Wittgenstein does not change _this idea_. But the problem is not that we cannot know a thing about the inner experiences of other people. The problem is that we cannot get from a inner experience to reality. We cannot cover the distance betwen the epistemic subject with her private experiences to the world. Here Wittgenstein has something important to say: we are already in the world; we act in the world and the success of our actions has our knowledge as an assumption (has the foundations of our knowledge, the hinge propositions, as assumptions). So we have to know the world (see OC).


273: What am i to say about the word "red"? [...]

var1: when I say that something is red I talk about red things; I could also talk about my sensation of red, but for this I need "another word";

var2: "red" means the property of being red (which red things have) and also my sensation of red (and another sensation of red for each person)

274: refers to vs. means

275: We can only refer to colours as public properties

277: But how is even possible for us to be tempted to think that we use a word to _mean_ at one time the colour known to everyone - and at another the 'visual impression' which _I_ am getting _now_? (see 39: "But why does it occur to one to want to make precisely this word into a name, when it evidently is not a name?"; here, as there, he seems to have Russell and/or his younger self in view; but these are not arguments; explaining a mistake is part of the therapy, not part of an argument showing why one says it is a mistake)

"When I mean the colour impression that (as I should like to say) belongs to me alone I immerse myself in colour - rather like when I 'cannot get my fill of a colour'. Hence it is easier to produce this experience when one is looking at a bright colour, or at an impressive colour-scheme."

Ok, but what was the reason to talk about my sensation of colour? And why would I "detach the colour-impression from the object" precisely when the object is the most intensely coloured? If the object is intensely coloured, the colour should "stay there". 

The rainbow is not vividly coloured and it is not a spatio-temporal object. When I see a rainbow a see an impressive colour-scheme (perhaps), but the point is that the rainbow is "not there". It does not become bigger as I move "closer to it". I cannot pass through it, as it were a real arch in the sky (see the French expression for rainbow).

"I see something, but there is nothing there. Then it cannot be said that I see something. I could say, instead, that I have some visual sensations. My sensations are produced by some objects which I can see (as is the case with the grid illusion, or even by some objects which I cannot see directly (like tiny bits of water etc.)." 

There is also the distinction between primary and secondary qualities. See Tractatus: Objects are colourless. "The the colours which I see are not real. I am just having colour sensations (but I do not say that "I see colour sensations").

278: "I know how the colour green looks to _me_." - I have a green qualia.
279: Imagine someone saying: "But I know how tall I am!" and laying his hand on top of his head to prove it.

It does not make sense to talk about tallness if we cannot compare our own height with the height of others. Similarly, it does not make sense to talk about "how colour green looks to me" if we cannot compare this "looks to me" with the "looks to" in the case of other people. 

Another use: "I know how tall I am", putting my hand on top of my mead means that I have my body at my disposal and  I could measure myself whenever I wanted to. But here the analogy breaks.

Another use: If we disregard measurement (as an attempt to quantify my height), it is as if I could "feel" my height (when I put my hand on top of my head). So this should prove that I know it (I have a qualia about it). But here the analogy also breaks, since there is not qualia for "having the qualia of seing red" (and if it was, how could I know that I have it?).

Another use: Someone challenges me in a way: "You have to bend your head when passing through that door." I place my hand on top of my head and reply: "But I know how tall I am". What I mean is that I see the door and I can estimate on sight if I could fit through it (without lowering my head, bending my back etc.) or not. Suppose someone challenges my ability to recognize a colour (or a particular shade of colour to go with the colour of another thing etc.). I can plausibly argue (convincingly or not) that I have that ability without really talking as if I was refering to my colour sensations. Here "I know how green looks to me" means "I have learned to identify green, to compare colours etc.". There is no emphasis on "me", as in "I do not know how it looks for you, but I know how it looks to me. And if there is, it's different: "I do not know about you, but I can manage to fit the green of the carpet I want to buy with this sample of green".

282: When we say that dolls feel pain we pretend that dolls feel pain. This use is "a secondary one", but not because we have the qualia of feeling pain (see the intentional stance). We naturaly attribute pain to living beings and by pretension we attribute it to inanimate objects (which we consider similar to living beings). 

283: What gives us _so much as the idea_ that living beings, things, can feel?

If we cannot use qualia here, how could we justify this claim?

The scenario of turning into stone is interesting. Contrast: I imagine that a whitch has a pain and she turns hereself into a cup of tea and she still has pains - I still see the cup of tee as a witch (but I could also imagine that the witch's pains were turned into something else, so I see the cup of tee as a witch, but she is not in pain anymore) / on the other hand, I imagine that I turn into stone - I can convince myself that this has happened so I stay still, I think I am a stone, but I'm still me and I feel some pain; I could also think that a stone cannot feel pain, so the pain is still present, but it does not have a bearer anymore (in a sense, I could say that I have turned into pain, or that my body turned into stone and my soul turned into pain).

"What has a soul, or pain, to do with a stone?"

"Only of what behaves like a human being can one say that it has pains." 

But if I close my eyes and imagine I am a stone I don't know if I behave like a human being anymore. Or perhaps I should say that I do not behave like a human being anymore and that's for sure.

284: "How could one so much as get the idea of ascribing a _sensation_ to a _thing_?"

This could also be understand in relation to 312 (pain-patches on the leaf).

285: "Think, too, how one can imitate a man's face without seeing one's own in a mirror."

What is this about? (it is somehow related to the talk of a transition from quantity to quality at 284, but I cannot say how right now.

286: [..] one does not confort the hand, but the sufferer: one looks into his face. 

Yes, but we do not confort the face. 

287: How am I filled with pity _for this man_?

I do not pity a body. Also I do not talk to a body, but to another person. What if someone would tell me 'I am not talking to you right now, but to your brain.'?

"I am not affraid of you, but of your gun." - Am I, really? Well, I could be affraid that the gun is old and could fire by itself and so on. I could also be affraid of a storm in a similar way. But if we are in an ordinary case, I am affraid of you. This could also be reformulated as "I am affraid that you could shoot me". What have we gained in this way? I do not talk anymore as if a person was the object of my fear. Now my fear is a propositional attitude (fear that...). But the object of my propositional attitude still contains a reference to you. "You could shoot me." - Now, what is the sentence talking about? A body? "But sentences do not literaly talk about anything." - Ok, what would a person uttering that sentence refer to (in the subject of the sentence)? 

*

When I was in my first grade I was already talking to other kids. I remember interacting with them, not only with their bodies. They were not just some factors from my environment. It was more than that, altough I do not feel the need to point out some particular experience that it was more than that. I could also say: 'If it weren't more than that I would have felt quite alone, but I didn't'. Now, suppose someone was to point out to me that there is a problem with the expression 'the other kids'. I imagine the following scenario. The person could have asked me who are the other kids and I would have pointed out to them and replied 'They are the other kids'. Then my interlocutor could have said: 'You are pointing to some bodies. Are these bodies the other kids?'. I can focus my attention only on someone's body, but after a while my attention would shift to that person. How does such a shift happen? I look at a chair. Then I force myself to shift my attention to it "as a person", as if there was a person standing next to me. Now it is as if the chair might expect something from me. I do not feel pressed to talk to the chair, but it is as if I could do it if I wanted to. Ok, but am I supposed to say that "the other people" are some sort of grammatical mistake? "There is nothing like a person. Just my attitude towards my environment (or something from my environment) has changed." - But before being affraid of a storm as a natural phenomenon human beings were affraid of a person who produced the storm (or of a strong wind conceived _as a person_). Imagine a painting: every object has eyes, a facial expression and a mouth. Is that the world of a primitive human? Historical accuracy aside, what if someone viewed the world like this at first and then started to regard some parts of her surroundings as inanimate objects. Now the problem would be: "How could she do that? If everything she saw was a person, how could she come to see something as 'not-a-person'?" Now suppose she were to say that she sits on a chair and a different kind of interlocutor was to answer: 'Isn't it that the chair is holding you in her arms? What is it that which you sit uppon? Point it out to me. [...] And now you are pointing at a person. How could there be a non-person ("a mere object") in there?'. If she replied that the chair as a mere object was more like a dead person, the interlocutor could say: 'But dead persons are still persons, only that they do not act. They might sleep, but they are still persons'. And then she might get convinced that "a mere object" is some sort of grammatical fiction. This is an awkard scenario, but not an impossible one.

The problem, however, remains the same. We distinguish between interacting with objects and interacting with persons. 

When we try to establish whether or not a person is the same with someone from the past we use some criteria. Such criteria are objects, so we have something to do with objects. Even if we reached the point where we could say what the ultimate criteria are, the question 'What is a person?' would be left untouched. Isn't _this_ weird?

I play a MMORPG. Now think about my attitude towards a NPC and my attitude towards (the avatar of) another player. What is the difference. I interact with NPCs as if they were people. I interact with (the avatars of) other players as if they were characters in the game. Both cases involve some sort of pretension, but I pretend to do different things. In the case of other players I pretend that they are "in the game", while in the case of NPCs I do not have to pretend that. I might, however, treat whatever is "in the game" and behaves like a person as a person, so the difference in pretension might not be relevant, after all. Were a robot to act very much like a person, I would treat it as a person. And it would be a person - according to Wittgenstein and Turing, I think. When I try to think of a chair as if it was a person I change my talk ("the chair sits close to me" vs. "there is a chair in this room"). Yes, but shouldn't this change also involve attributing some qualia to the chair or to the robot? And if not, why not? Eike von Savigny: "Ok, I am a zombie. No problem." - I do not need to attribute some qualia to myself in order to consider myself a person. 

"But you already are a person. How could you doubt that you are?" - What if I discover that I am a robot? (as Astro Boy does, at some point) "That scenario might make you realize that no 'experience which you have' does really prove that you are a person, but you would still believe that you are a person. (If someone put you into a rigid armor and control it to make you kill thousands of people you would still feel bad about it. This is somehow similar)" - But then all I need to do is to attribute the same to the robot, to think that she could not seriously doubt that she is a person. "But you cannot doubt something which you cannot understand. Of course the robot cannot doubt that she is a person, but not because she believes that she is a person and cannot get rid of that belief. The robot just does not understand what it is to be a person, so she cannot doubt it." - Well, if we ask the question 'What is a person?' seriously, then perhaps we too do not know what it is to be a person.

"Ok, so you say that the distinction between persons and objects is arbitrary?" - Perhaps the distinctions comes down to separating what resembles us from what does not resemble us. Or what can occupy in our practices a position similar to the one occupied by us and what cannot.

"So you are a naturalist, after all". - I am not an anti-naturalist, in any case. I would never say that is wrong to talk about everything arround us in a naturalist vocabulary. What I say is that it would not be practical to talk only like this, even if it was possible. It might be enough to talk only like this when facing any theoretical concerns, but there are other concerns too. What I am inclined to believe is that our talk of persons, actions, intentions, responsibility, reasons and so on is extremely practical when it comes to those other (more frequent and perhaps more important) concerns, while talk about sensations, feelings and other psychological entities and processes is only misleading.

"And how can you be so sure that you can distinguish between the last two vocabularies?" - Well, this is not about two sets words, but about two different attitudes. The psychological vocabulary is used while mimicking the attitude of the scientist, but is used to talk about persons and their interactions. I think if we do really want to talk of humans from a scientific point of view we could talk about human bodies, human brains and human behaviour. That should be enough. In this sense, "does she love?" me could only mean "is she going to tell me that she loves me or react in this and that way if I told her that I love her?" or perhaps "is this and that happening in her brain?" The "qualia of love in her psyche" is the cog which is not connected to anything else (and which nobody but her could ever know anything about). I might still be tempted to think about "her feelings", but not because I am really curious to know something about her feelings. The problem is that I have the feeling that something is left out by the scientific version of my initial question ("does she love me?"). What is left out, however, has nothing to do with any theoretical concerns. I aim to establish a certain personal relation with her - that of lovers (or to maintain that relation etc.). My concern is whether to invite her to take up a role in that relation or not. Taking up such a role is a responsibility, so I might be wondering whether it is right for her to assume that responsibility of for me to ask her to assume it and so on. Is it ok for me to expect (or to ask) her to assume the role of a lover towards me? It could be that such question might be somehow translated into "scientific" ones by someone who is theoretically inclined, but my interest in asking them is completely different. If I could learn to express such interests in scientific talk, perhaps everything would be ok, but then one could doubt that that talk should be called scientific talk anymore. In any case, it would be pointless to answer a question like "What are you refering to when you are wondering whether you and her should be lovers or not?". I am not talking about my body and her body, my mind and her mind, my consciousness and her consciousness, and not even about my soul and her soul. Am talking about me and her being lovers, not about some souls (mine and hers) being lovers. There is no ontological commitment at stake here. I am talking about a completely different kind of commitment - commitment to a kind of personal relationship, so why would any one bring ontological commitment into discussion?

"Right, but whose commitment are you talking about." - Mine and hers. Now, if you still have theoretical problems about _this reply_, please go to another room and talk to a neuroscientist.

"But isn't this just refusal to think about a philosophical problem?" - Well, it's more like refusal to consider a joke told at a funeral funny.

"Ok, but you are not really considering whether you and another person should be lovers _right now_". - Right, but there is nothing to be achieved by imagining you are at a funeral and then pretending to tell a joke in that context. 

*

2.10

Different uses of a mark on an object: to enable one to recognise an object, to enable one to remember the object easier (as a mnemonic device), as a tag on a box, as some visual aid when counting a range of objects ("I have counted up to this one") and also as a visual aid for an entire range of activities - I mark the place where I want to make a hole in an object, I mark your height on the door frame etc. 

*

Directing my attention towards an object. Why is this different when the object is a public one? Because what I do is integrated into a series of practices - pointing to things, grasping things, showing them to other people, saying "No, look at this" or "You are not paying attention to what I am showing you" a. s. o. 

Directing my attention to some private object is disconnected from all these practices. Perhaps one could talk about "paying attention to a pain which I cannot ignore" or "some strong sensation which catches my attention" etc. There is, however, a difference between these two cases - directing your attention towards something vs. having your attention grasped by some strong sensation (like a loud noise or a strong colour). 

I hear a loud bang which grasps my attention. I make a mark, planing to make a similar mark whenever I hear a similar loud bang. The mark is not a name referring to my sensation yet. Also, I could mark the physical bang (a physical event) or my hearing sensation. Now, if I disregarded what I believe about how the sensation was produced, it is hard to see why would I call it a sensation (and not a feeling, for instance) and also how could I re-identify it. Also, the sensation cannor bear the mark on itself. At most, I can try yo fix a connection between the bang and the mark "in my mind" . 

[...] 

Wittgenstein: private experiences are semantically irrelevant, because I cannot refer to them in a language only I can understand. They are epistemically irrelevant because they are semantically irrelevant. 

Sellars: One could somehow manage to talk about these private experiences in a public language ("F looks red to S.") and consider them the empirical datum - the foundation of all our knowledge about nature. 

Right, but why does the empiricist want to save the idea of an empirical datum? Because she wants to answer the skeptical challenge. Wittgenstein has his own answer to that challenge in On Certainty. 

My understanding: 
(Chair) There is a chair in this room. 

Empiricist justification for (Chair): I have chair-like sensations in this room, good reasons to believe this is not a hallucination etc. 
Sellarsian empiricist: There is something looking like a chair to me in this room etc. 
Wittgenstein (OC): The question whether there is a chair in this room or not does not rise in any usual circumstances. A situation could be imagined in which we would want to test for the presence of a chair in this room. In regular cases, however, we do not know that there is a chair in this room (we do not play the game of knowledge with respect to the chair in our room), but we are certain that there is a chair in the room, since this is an assumption of our actions involving the chair (a necessary condition of us performing those actions as actions involving the chair);

Wittgenstein acknowledges, however, the idea that having the belief that (Chair) does not have to do with being in a psychological state of belief towards the content of (Chair). Since (Chair) could be epistemically relevant (we could perhaps justify something within the game of knowledge by it etc.) and we can be said to believe that (Chair) [...], perhaps there is a sense in which it could be said that we know that (Chair). Now, if we accept that A is a necessary condition for B IFF B is a sufficient condition for A (this could be disputed, of course), then from 
(Nec) Our knowledge that (Chair) is a necessary condition for our actions involving the chair in this room (under such descriptions). 

it follows that

(Suff) Our actions involving the chair in this room (under such descriptions) are sufficient conditions for our knowledge that (Chair). 

One could also want to talk about successful actions etc. 

This could also give us some empiricist moral: Our experiences are, in a sense, the basis of our knowledge, but only if we conceive them as (constituted by) actions performed by us in our environment. 

3.10

We can have a concept only insofar as we have a practice. This is a side moral of the RFC. No concept without a practice. One could, of course, imagine a non-existent practice (Wittgenstein does that a lot), but at least one has to have taken part to practices which involve imagining things (some sort of pretense) and to practices which involve the recognition of other practices (as it happens in anthropology or sociology or some of their rudimentary forms) in order to do that. 

Directing your attention towards X, taking something to be a sensation, ownership, naming, believing etc. - all these concepts presuppose at least a practice. Private objects, by themselves, cannot produce a practice. This is the link of the PLA to the RFC. 

Robinson can invent practices, while the Cartesian subject cannot. That is why further correct application of the ostensive definition are not necessary in Robinson's case. But they are necessary in the case of the Cartesian subject. In that case, however, since the correctness of the future cases depends on the correctness of the ostensive definition, we have a vicious circle. This is a better form of Wittgenstein's PLA.