The example that continually comes into my head, unbidden, in this discussion is Philip K Dick's novel 'The Man in the High Castle', which is a counterfactual in which the Axis powers won WW2. Note the neutral manner in which that sentence describes it.

If asked to tell somebody what the book is about I think I would say 'It imagines a world in which the Axis powers won WW2'. Based on recent posts above, it appears that Kripke might say 'It supposes that the Axis powers rather than the Allies won WW2'.

I can't identify a tangible difference between the two. To me they just look like different strings of words that gesture towards the same concept - a vague concept but one that has enough solidity in it for us to enjoy the novel (admiring Dick's imagination and doing our best to overlook his execrable dialogue).

Historically, it was thought hat descriptions explained reference. Drawing attention to the falsity of this picture of reference was good for philosophy generally. A step away from armchair theorising, towards looking around. A continuation of Wittgenstein and Austin, to my odd way of viewing things.

One does not need definite descriptions in order to refer.

One can use them, but they are not the essential, and not the explanation of reference.

Someone should turn this thread into a book.

One can use them, but they are not the essential, and not the explanation of reference. — Banno

I still disagree with this, because I remain unconvinced that anyone can refer to anything without (implicitly, at least) relying on the knowledge of the world that comes either from direct observation or description. It is arguable that even referring to what is directly observed would be impossible without descriptions (conceptions) that have already become internalized through enculturation.

When you refer to Aristotle, to return to the example, how did you acquire your idea of just who it is you are referring to?

I've been feeling a bit bad that what started as a reading group is mostly debate, and I'm partly at fault for that. The debate has been so interesting that I have not been able to resist the temptation to jump in now and again, even though my better angel says I should stay out and let the people that are currently working through the book have clear air.

I wonder would it be better to create a new thread that discusses the book, and refers to the reading group thread, and leave the reading group thread itself for those that are currently working through the book. We used to have that approach with debates, where there was a formal debate thread for the two participants (in this case it would be the official participants in the reading group) and a discussion thread where anybody could discuss the debate and issues arising in it.

I think it would be way too difficult, given the length of the thread, to dissect the existing thread into reading group and debate. But we could decide that future debate between Kripke sceptics and Kripke enthusiasts (as opposed to debate between Kripke enthusiasts) should go in the other thread.

It's just a suggestion, and I'm sure that those of us that are Kripke sceptics would be happy to stay out of this one if that's what is preferred (although I'd still read it because I find it very engaging).

If asked to tell somebody what the book is about I think I would say 'It imagines a world in which the Axis powers won WW2'. Based on recent posts above, it appears that Kripke might say 'It supposes that the Axis powers rather than the Allies won WW2'. — andrewk

If one imagines a possible world at which the Axis powers won WW2, then one can imagine it such that Hitler lived to be 72. But when one evaluates the counterfactual conditional statement "If the Axis powers had won WW2, then, Hitler would have lived to be 72", the truth value of this counterfactual conditional proposition doesn't appear to be tied to what it is that we can imagine. It rather depends on what we are entitled to believe would necessarily have happened if the counterfactual antecedent had been true. There are difficult issues regarding what it is, besides what is stated in the conditional antecedent, that is taken to remain the same, or vary, with respect to the actual course of history, while evaluating the necessity of the consequent. (Can we allow for backtracking?) In my view, those issues are settled by pragmatic considerations. In any case, the truth conditions of a counterfactual conditional proposition aren't dependent on what one might imagine, but rather on what one is entitled to expect would necessarily have occurred or would necessarily have been the case if the antecedent had been true.

I still disagree with this, because i remain unconvinced that anyone can refer to anything without (implicitly, at least) relying on the knowledge of the world that comes either from direct observation or description. — Janus

My emphasis.

The disjunction is your Get Out Of Gaol Free card.

Multiple arguments against the second disjunct were produced over the last twenty-odd pages. I'm not going to go over them again.

And Kripke pretty much agrees with the first disjunct, although he would phrase it in terms of cause rather than observation.

(Actually, I take that back; how could observation alone be adequate for individuation to occur? It strikes me now that individuation only occurs as part of a language game.)

For my part i am happy to intermix dialogue and exegesis.

This misses the point. Indeed, all of those particular items cannot exist without their elemental constituents. — creativesoul

I am usure what it is that you mean with the phrase "elemental constituent". Also, could you explain in which way knowledge of the elemental constituents of an individual can be made use of in order to refer uniquely to this individual?

I think it would be way too difficult, given the length of the thread, to dissect the existing thread into reading group and debate. — andrewk

I'm banking on your or another's willingness to take up that challenge. I have seen valuable insights over the last three or four pages as to the issue of identity and time tracking contra essentialism.

I'm interested in what you said in the first para that 'this is the sort of counterfactual consideration that we rely on when planning future actions or when we are gathering evidence for the existence of causal relations'. I would exclude considering future actions from that because that is usually a case not of imagining the past being different, but rather imagining more than one different possible future, neither of which contradicts current knowledge. — andrewk

It's true that in the course of practical reasoning one usually restricts the consideration of options to those that one has the power and opportunity to realize. However, the subjunctive conditional "If I will do X then Y will occur" could be true even if, unbeknownst to me, I lack the power or opportunity to do X. In that case, the antecedent is necessarily false conditionally on the actual present state of the world being what it is. But so long as we don't equate the actual with the necessary, the subjunctive conditional proposition might be still be true.

Those are not counterfactuals but rather considerations of future possibilities - I call them 'Hypotheticals'. By 'gathering evidence of causal relations' I assume you are referring to the attempt to develop scientific theories. I agree that counterfactuals can play a key role in that but it seems to me that they work perfectly well with my interpretation of counterfactual, and don't require a Kripkean interpretation.

Kripke's conception of counterfactual conditionals and of possible worlds is very deflationary, in Naming and Necessity. If I remember, he proposes to construe phrases such as "there is a possible world at which P" to be meaning no more and no less than we would ordinarily mean when we say "It is possible that P might have happened".

My view on these issues is set out in somewhat more detail in an essay I wrote a couple of years ago: Hypotheticals, Counterfactuals and Probability. Those are still my views. — andrewk

Thanks, I'll put this as high up in my reading list as feasible.

The example that continually comes into my head, unbidden, in this discussion is Philip K Dick's novel 'The Man in the High Castle', — andrewk

LOL. I brought up the same work earlier in this thread to make the same point. I think error and ignorance are the better arguments. You don't have to know any definite description of Hitler to refer to him, as when you ask who he is. Or you could refer to him, but be wrong in thinking he was an artist.

BTW you asked above if I was seeking to defend descriptivism. I think I probably am, but that doesn't mean I think it's the best theory. I see Wittgenstein's language game approach as the best explanation of language, including proper names. But despite its faults (which I think are different from those that Kripke claims) I think there's a lot of valuable insight in Russell's theory of descriptions, and I am unable to find any such value in Kripke's theory. — andrewk

Myself, I think Russell's theory of definite descriptions is basically right. But the correctness of this analysis doesn't entail descriptivism, unless one also holds that proper names can be analysed as definite descriptions. That's not something Russell himself believed, regarding "logical proper manes", which he took to only refer to direct object of acquaintance such as sense data and the referent of "I". Gareth Evans developed an account of proper names that is quite indebted to Kripke, to Wittgenstein, to Putnam, and to Russell, although he ditched Russell's antiquated Cartesian epistemological restrictions on direct objects of acquaintance. But he maintained a relaxed version of what he called Russell's Principle: that a person cannot be thinking about an object unless he knows, in some non-trivial way, which object he is thinking about.

But he maintained a relaxed version of what he called Russell's Principle: that a person cannot be thinking about an object unless he knows, in some non-trivial way, which object he is thinking about. — Pierre-Normand

Isn't that the principle or knowledge by/of acquaintance stated another way?

P.95. Another example, and one which shows nicely the methodological tool Kripke introduced, and which puts an end to descriptive theory. The tool is to look at a sentence that presents an erroneous description, and notice that nevertheless, it is about the thing misdescribed.

So Newton is described as "the man who first thought that there is a force pulling things towards the Earth". Despite being wrong, that sentence is about Newton. Further, it is about Newton, even if there is no other description available that can serve to single Newton out.

The joke about George Smith squaring the circle (p.95) serves to show that Kripke is not after an alternative theory of reference.

Isn't that the principle or knowledge by/of acquaintance stated another way? — Wallows

Quite right. As I said, Russell thought we could only be acquainted with sense data and with our own thinking 'selves'. Everything else, including the referents of most ordinary language proper names (of cities, human beings, etc.) only are knows by description. Evans agrees with Russell that descriptive content isn't a form of singular reference (but rather must be analysed into existentially quantified statements), but he argued that proper names and demonstratives can refer to much more than just sense data.

Notice that the preceding outline hardly eliminates the notion
of reference; on the contrary, it takes the notion of intending
to use the same reference as a given. There is also an appeal to
an initial baptism which is explained in terms either of fixing a
reference by a description, or ostension (if ostension is not to be
subsumed under the other category)." (Perhaps there are other
possibilities for initial baptisms.) Further, the George Smith
case casts some doubt as to the sufficiency of the conditions.
Even if the teacher does refer to his neighbor, is it clear that he
has passed on his reference to the pupils? Why shouldn't their
belief be about any other man named 'George Smith'? If he
says that Newton was hit by an apple, somehow his task of
transmitting a reference is easier, since he has communicated a
common misconception about Newton.

Reference takes place within a definite description, even if that description fails.

The deeper point is that even in order to set up a definite description, one must already be able to refer to the individual. Hence reference cannot presuppose descriptions.

Italics - this is important.

And of course this falls in directly with the discussion of Wittgenstein's dismissal of ostension as the basis for language - found in the Witti reading thread. Ostensive definitions presuppose an understanding of reference.

https://thephilosophyforum.com/discussion/comment/237129

And then Kripke moves on to Identity. Another controversial topic.

The deeper point is that even in order to set up a definite description, one must already be able to refer to the individual. Hence reference cannot presuppose descriptions.

Italics - this is important. — Banno

You don't have to know any definite description of Hitler to refer to him, as when you ask who he is. — frank

Frank's example may serve as a useful illustration. If someone asks about Hitler, and they do not know who he is (i.e. do not know any definite description of him) then they will probably be asking because the name has come up in some context, wherein others do know who Hitler is. How will people inform the asker as to who Hitler is, if not by description(s)?

So, of course the asker, although they do not know who Hitler is, can refer to Hitler (simply by saying the name in the question "Who is Hitler'?, but that reference is reliant on the other's knowledge about Hitler. Say no one knew who Hitler is, then who would the asker be referring to? You could say that they are all referring to some hypothetical character called 'Hitler', but without any descriptions at all of Hitler the whole thing would be utterly empty. What could they possibly say about this character Hitler that did not consist in a description however minimal? 'He was just a man' and that's it? A man, then as opposed to a woman, or a donkey?

@Banno, what does Kripke have to say about accessibility relations between possible worlds? Does he even go into this important topic?

I'm quite interested to get into the logical aspect of this book. If I can assert a possible world or stipulate something about it, then existentially that stipulation will always be dependent on this world.

But, if you stretch the concept to the extreme, then what about possible worlds where accessibility relations fail? And, more to the point about counterfactuals, then it would be tautologous to say that they some are necessary or possible.

When is a theory undone.

At what point is it unreasonable to keep believing in a theory.

It's commonly understood that any theory can be kept from rejection by the addition of suitable hypotheses, ad hoc. And that at some point one ought just drop this process and say "yeah, that theory is buggered".

The theory at hand is that for any given proper name there is always some definite description that picks out one individual, that individual being the referent of the proper name.

Look at the question "Who is Hitler?". Ask yourself "who is this question about?" Isn't there something quite infelicitous about claiming it is about anything other than Hitler?

Now if you don't see that, so be it. You can add whatever auxiliary hypotheses you wish in order to save your theory. Go ahead and make the question reliant on the other's knowledge of Hitler, but all you are saying is that language occurs in a community.

Quite a bit. If you are really interested, i would commend any of the innumerable introductions to formal modal logic. It's really outside of the scope of my replies here.

You don't have to know any definite description of Hitler to refer to him, as when you ask who he is. — frank

I wonder if one is referring to anything at all when one asks who somebody is. Such a question typically comes up when one has overheard a conversation that sounds like it is about someone. The temptation is to say that the eavesdropper is asking 'About whom are you talking?' (because while ignorant of history and current events, the eavesdropper is quite particular, in an old-fashioned way, about grammar).

I'm inclined to say 'not so fast'. I think what the eavesdropper really wants is to have the story explained to them in a way they can understand.

Consider the timeless mondegreen: The minstrel is overheard telling his fans about how when Robin Hood died, they took him to the forest and they laid him on the green.

The eavesdropper hears:

'They took him to the forest and the Lady Mondegreen'

and asks 'who is the Lady Mondegreen?'.

To whom is the eavesdropper referring? Are they referring to the non-existent Lady Mondegreen?

I say that question is the wrong question, and writing dissertations about it misses the point. The right question is 'what does the eavesdropper want?' The answer is that she wants to understand the story that was being told.

I say that question is the wrong question, and writing dissertations about it misses the point. The right question is 'what does the eavesdropper want?' The answer is that she wants to understand the story that was being told. — andrewk

That may be the case, but then the eavesdropper could understand the story (well enough) without knowing who the person talked about is. So, the fact that supplying her with a definite description of the peson holding the name is sufficient for that purpose (i.e. understanding the story well enough) isn't sufficient for showing that the description determines the meaning of the proper name.

Furthermore, Kripke would readily grant that providing a definite description often is sufficient for fixing the reference of the name (as used thereafter by the eavesdropper), even though the description doesn't determine this reference all by itself (and hence isn't semantically equivalent to it). That's possible because of the social character of meaning. Supplying a definite description can be a means of co-opting a new user into an already existing name-using practice. This new user, when she has had the meaning of the proper name conveyed to her by means of a definite description, is thereafter enabled to think (and talk) about this named individual thanks to there already being other participants in this practice who are directly acquainted (or 'causally linked', Kripke would say) with the named individual.

I feel that my approach is the most natural in the world. — andrewk

That's not really relevant. There's a body of data w.r.t. how counterfactuals behave, and a theory to capture them. What you feel about the naturalness of anything doesn't matter.

Myself, I think Russell's theory of definite descriptions is basically right. — Pierre-Normand

Russell's theory is probably not right. It makes a number of wrong predictions as to the behavior of definite descriptions in embedded environments.

https://www.readcube.com/articles/10.1007%2Fs10988-010-9072-3?author_access_token=SVx2QIwFk3kTTo8IyEVDRve4RwlQNchNByi7wbcMAY7iNlNgpGRzM_WD4_srx4OSluaaYmnLCu53v9bvVDa92yUqe-uN_ec2lwOPjxmkaterOX7qM5JgxBKBMjrqfPx6CudiP6v1X4Hp3jXoWtKFow==