** متابعات ثقافية متميزة ** Blogs al ssadh
هل تريد التفاعل مع هذه المساهمة؟ كل ما عليك هو إنشاء حساب جديد ببضع خطوات أو تسجيل الدخول للمتابعة.

** متابعات ثقافية متميزة ** Blogs al ssadh

موقع للمتابعة الثقافية العامة
 
الرئيسيةالأحداثالمنشوراتأحدث الصورالتسجيلدخول



مدونات الصدح ترحب بكم وتتمنى لك جولة ممتازة

وتدعوكم الى دعمها بالتسجيل والمشاركة

عدد زوار مدونات الصدح

إرسال موضوع جديد   إرسال مساهمة في موضوع
 

 Rigidity for Temporal and other Non-Alethic Modal Talk

اذهب الى الأسفل 
كاتب الموضوعرسالة
free men
فريق العمـــــل *****
free men


التوقيع : رئيس ومنسق القسم الفكري

عدد الرسائل : 1500

الموقع : center d enfer
تاريخ التسجيل : 26/10/2009
وســــــــــام النشــــــــــــــاط : 6

Rigidity for Temporal and other Non-Alethic Modal Talk Empty
12032016
مُساهمةRigidity for Temporal and other Non-Alethic Modal Talk

Just as rigidity might have bearing on different grammatical categories, not merely the paradigms to which Kripke applies it, rigidity might also have bearing on different modalities. For example, just as ‘the brightest non-lunar object in the evening sky’ seems to designate Venus with respect to the actual world but not with respect to other possible worlds in which Venus is obscured by some other body like Mars, just so the expression seems to designate Venus with respect to the present time but not with respect to other times at which Venus is obscured by some other body. By contrast, ‘Hesperus’ and ‘Phosphorus’ seem to designate Venus with respect to all times. (For cautions about the analogy, see Gómez-Torrente 2006, p. 249.) Therefore, again, ‘Hesperus = Phosphorus’ seems to be true regardless of the time of utterance; ‘Hesperus = the brightest nonlunar body in the evening sky’, on the other hand, is false when uttered as a passing comet outshines Venus.
Whether we are talking about alethic rigidity (designation of the same individual with respect to other possible worlds) or temporal rigidity (designation of the same individual with respect to other times), a commitment to rigidity will be attended by similar sorts of metaphysical and epistemic commitments. Here it will do to indicate briefly some ontological parallels having to do with the status of individuals with respect to other possible worlds, on the one hand, and times on the other. Just as recognition of alethically rigid designation carries with it a commitment to some sort of transworld identity (discussed later), so the recognition of temporally rigid designation carries with it a commitment to some sort of “transtime” or transtemporal identity. Similar alternatives to rigidity are available to theorists who would deny that the relevant designators are indeed rigid—and the choices are similarly motivated. Hence, if we adopt the ontological position that individuals are instantaneous so that we find it natural to think in terms of the “stages” of you since, although the stages are not identical, they're similar and ordered in relation to one another spatiotemporally or causally, then we might naturally take names like yours to range over different instantaneously-lasting individuals, and so conclude with Varzi (2003, p. 387), that “the rigidity claim can't be right and a counterpart-theoretic semantics seems required.”
Familiar, broadly applicable reservations about the reduction of modal talk to counterparts talk arise in their own turn: namely reservations according to which ordinary speakers' accurate talk of a future and a past for you uses ‘you’ and your name rigidly,[3] since ordinary speakers presuppose or assert that something in the past and something in the future is identical to you. The counterpart substitute, it will be said, is either a misdiagnosis or a tacit concession to the falsity of ordinary claims and a masking over of that falsity by means of a so-called reductive analysis. Quinn, for example, anticipates the dilemma, maintaining that identity through time is what is needed to honor claims to an individual's lasting through a day, and maintaining further that if we conclude that we don't really last over time, and decide anyway that “it would be convenient to speak” in some counterpart-reductive fashion in order to be able still to assent when ordinary speakers innocently say ‘You were around yesterday and will probably be around tomorrow’, then we end up masking our disagreement with what ordinary speakers literally claim (1978, p. 348). Metaphysical positions of great consequence are thereby concealed: ‘you'll be around tomorrow’ is now, according to the newly proposed reductive reading, to be accepted on the liberal terms that tomorrow there is a line of “momentary creatures endowed with similar properties,” none of whom is really you. But were a reductionist to affirm correctly the usual sentence, now eviscerated of the usual content, the affirmation should bring you “no comfort at all” (1978, p. 350; see also p. 347). The hope is that you yourself are around tomorrow; if others are around too, that might be good, but it is not enough to satisfy your hope. (So we would put it if we believed in the reality of different times. More neutrally, we might put the point as follows. When we conceive of tomorrow, you're part of what we're conceiving as present, hopefully; if others are around too, that is irrelevant to this hope.) Since the sentence ‘you'll be around tomorrow’ on its original interpretation expresses comfortingly that your hope is to be realized, and since the reductive assent does not, the reductive assent hardly ends up conveying the appropriate opinion about what ordinary speakers are talking about.
Kripke presents rigidity first as an alethic notion but there is also a temporal interpretation, as we've seen. There are other interpretations too, which is not surprising if rigidity is fundamental to all our modal talk: and it seems reasonable to say, with Føllesdal, that it is. “All our talk about change, about causation, ethics and knowledge and belief, as well as about the other modalities, presupposes that we can keep our singular terms referring to the same objects. To the extent that we fail, these notions become incoherent” (Føllesdal 1986, p. 111).
Furthermore, a term might rigidly designate under one interpretation of rigidity but non-rigidly designate under a different interpretation of rigidity: thus, ‘Hesperus = Phosphorus’ is apparently not epistemically necessary, even though it seems alethically and temporally necessary; accordingly, ‘Hesperus’ and ‘Phosphorus’ are plausibly thought not to be epistemically rigid. Indeed, as Fitting and Mendelsohn (1998, p. 219) observe, “it is hard to see whether there are any rigid designators under the epistemic reading.” This entry will address an alethic interpretation of rigidity for the most part.

2. Relations Between Rigidity and Associated Theories

2.1 Relations Between Rigidity and Transworld Identity

A rigid designator designates by picking out its designatum not just with respect to the actual world—as things are—but with respect to all other possible worlds in which it is present, as well. That a proper name like your name is rigid by way of picking out its same designatum (you) with respect to possible worlds or situations different than those actually obtaining, though that claim is not without detractors, is a claim that enjoys great appeal. Call the relevant claim “transworld identity.” The claim is associated with a host of others that might enjoy much less appeal. Many such claims turn out to be separable from rigidity. Nevertheless, transworld identity remains a substantive commitment of rigidity, at least as classically understood.

2.1.1 Questionable Controversy

The commitment to transworld identity should not be confused with a commitment to realismabout possible worlds. Prominent champions of rigidity seem favorably inclined to realism (see Davidson's 2003 collection of Plantinga's essays; see also Kripke 1980, p. 16). But there is no reason an instrumentalist about possible worlds like Bas Van Fraassen, say, who regards possible-worlds talk as a useful fiction, cannot appeal to rigidity (as Van Fraassen does: 1984). What matters to rigidity, or rather to the position that a designator for an individual x is a rigid designator, is that in using that designator to entertain the various counterfactual situations or worlds S in which x exists, whether we do the counterfactual entertaining about x by way of apprehending abstracta, say (if worlds are real objects, abstract ones), or by way of pretending (if counterfactual talk is a fictional pretense, say, because worlds are not real) or whatnot, we entertain situations (or pretend to do so) that, were they to obtain, would contain x: so we tacitly commit to “the impossibility that S obtain and x fail to exist” (Plantinga 1974, p. 96). Accordingly, suppose I say to you, “If you'd driven to the park instead of walked, you'd have missed the refreshing breeze.” Here I entertain a nonfactual possibility associated with your driving to the park. Whom, in that entertained possibility, do we think of and speak of, as driving and not walking? The transworld-identity theorist says: you. No one else “gets into the act” as the driver according to the transworld-identity theorist (as Lewis would frame that view, without sympathy: 1986, p. 196).
Realism about worlds, then, is not something rigidity or transworld identity comes attached to; non-realists have their version.
How about essentialism? Whether we want to say that rigidity or transworld identity is committed on that score might well depend on the form of essentialism in question. Not many philosophers would dispute Cartwright's observation that “‘Shakespeare’ is rigid only if the manShakespeare could not have existed without being Shakespeare—or as we might equally well say, only if it is essential to Shakespeare that he is Shakespeare” (Cartwright 1998, p. 69). So if Shakespeare could have been Obama instead of Shakespeare, ‘Shakespeare’ would not be rigid.
But even if appeals to the rigidity of a name like ‘Shakespeare’ (or related appeals to the transworld identity of Shakespeare) cannot be divorced from essentialist commitments like the forgoing, such appeals will not commit us to much at all about what a designatum is essentiallylike with respect to the various possible worlds in which it exists to serve as the designatum: whether Shakespeare could have been a brute, say, or had different parents (matters over which Kripke tellingly disagrees with, say, Plantinga: Kripke 1980, pp. 110ff; Plantinga 1974, pp. 65ff).
It is far from clear then, that transworld identity and the rigidity that depends on it come attached to substantive essentialism of a sort that settles salient disputes in philosophy about which of your properties are essential to you. Closely related to the issue of substantive essences is the issue of substantive criteria, perhaps qualitative criteria. Such concerns, as developed in the metaphysically oriented direction of transworld identity, will be addressed broadly speaking in connection with counterpart theory; here let us move in an epistemological direction in order to address concerns about what we might call transworld “identification.”
Kripke addresses the objection that we cannot meaningfully talk about you, with respect to another possible world, without first having some qualitative criterion of identity, some qualitatively distinguishing mark that allows us to pick you out from other objects in the world at issue, in order to assign your name to the right person, i.e., to you, as the individual that satisfies the qualitative criterion. This criterion would appeal to your essence (or be “an essence”[4]; bear in mind, for this example, the minimal requirements of weak necessity from 1.2), in the minimal respect that the criterion must be something that you and you alone have with respect to any given possible world. As an objection, the worry is that we know of no such qualitative criterion so we can not meaningfully discuss you, with respect to any merely possible world.
One way to defend rigid designation in view of the forgoing objection would be to insist that as successful speakers, we do after all know of qualitative criteria by which we are able to distinguish individuals like you by name: some historical figures have in effect adopted this position by way of defending rigid designation.[5] But Kripke would concede that we speakers are not privy to the content of any qualitative criterion characterizing you. It is hard to gainsay the concession. We would certainly be hard-pressed to say in any enlightening way just what qualities characterize you with respect to the various possible worlds. Kripke would evidently agree that there is something that it is to amount to you, as opposed to anything else, and that your essence is what qualifies an object—you—to be the designatum of your name with respect to any possible world (he doesn't say that your essence is qualitative though, and he outright rejects certain versions of that claim). But Kripke would deny that a speaker would have to know this essence, or to rely upon any other nontrivial criteria distinguishing you from world to world, in order to refer to you by way of a rigid designator (Kripke 1980, pp. 15-20; see also Plantinga 1974, pp. 93-98).
Still, we can imagine why others have demurred. After all, if we lack the ability to say in an informative way just what it is to amount to you, as opposed to anyone else, with respect to counterfactual situations, then how can we as speakers know enough of what we're talking about in order to select you as the designatum, as opposed to anyone else, with respect to counterfactual situations? How can we know that you, no other, are the subject of our thought and talk? We certainly cannot define your name rigidly by means of essence-invoking criteria if we don't know such criteria.
In response, Kripke (1980, p. 44; see also Plantinga 1974, p. 97) insists that we may give a satisfactory account of how we designate you, trans-world, without giving a satisfying account of your essence or of criteria distinguishing you. Whatever the content of your essence, we may simply stipulate that the bearer of your name is to be you when we consider other possible worlds in which you do different things than those that you in fact do. Accordingly, we may take for granted that you are the protagonist of the different actions at those different worlds, despite our not thinking about what your essence is.
There would appear to be something robustly sensible in this suggestion. It seems sensible to suppose that ordinary speakers do something like stipulate: a speaker might naturally insist, “I'm thinking of you driving instead of walking, by George!—I have told you that. What's the difficulty?” It also seems sensible, at least on first blush, to take the success of this stipulation for granted, at least for many purposes, in many contexts.
Even so, the worries that motivate an appeal to stipulation still remain, in large part, to be accounted for, after we have provisionally set them aside by approving the appeal: the appeal to stipulation is more like a promissory note than the satisfaction of an explanatory obligation. The appeal to stipulation puts off for another occasion any attempt to resolve how we succeed at doing what we take for granted that we manage somehow to do: namely, how we succeed at referring to the right individual, by means of our stipulative effort. There has to be some “reasonthe stipulated situation, when we use a name, contains the object it does” (Sidelle 1995, p. 99n.4) rather than likely competitors. It is hardly obvious what that reason would be. To see why, consider that in order successfully to stipulate that a name is to follow just you, as a rigid and therefore transworld tracking device, our stipulative effort has to be able, across worlds, to allow us to distinguish what is you from what is not you but is instead your body (say: assume you are not your body). How is this to be done without specifying criteria, if you were with your body when your parents smiled in your direction and baptized you with a rigid designator, saying “We have decided on a name for the birth certificate: …,” thereby stipulating that you are to be called by the name they chose for you? “It is not by magic,” as Jackson (1998, p. 82) reminds us, that your name “picks out what it does pick out” rigidly—namely you—despite the competition against you presented by a different candidate for designation—your copresent body.
So how does it happen that the stipulative effort attaches your name to you and not competitors? According to a Lockean line of thinking, the indispensability of criteria becomes clear when we consider that question; stipulation is no substitute for criteria. For John Locke, unless a speaker baptizing a term were to have essential criteria in mind to demarcate the real referent—you, in this instance—from alternative candidates like your body, the speaker would stipulate a rigid designator in vain: his names would turn out, disappointingly, to be “the Signs of he knows not what, which is in Truth to be the Signs of nothing” (Essay III.ii.2, p. 406). On this line of thinking, a stipulation leaves too much indeterminacy to secure the right rigidity. To insist that the right designatum is secured without criteria seems, according to this line of thinking, to insist on stipulation getting it right by magic.
So there is more to be said about transworld criteria, even if an appeal to stipulation is all right so far as it goes. But because the relevant contributions to the discussion become scattered and elliptic, a supplementary document is handy for elaboration:
اقتباس :
Stipulating Identity Trans-world, Without Qualitative Criteria for a Designatum to Satisfy

2.1.2 Residual Metaphysical Controversy

It is not obvious that rigidity and its attending transworld identity run up against anti-realism, nor against any substantive essentialism and its gainsayers, nor against the intuition that there must besome way to assign reference that avoids (perhaps by way of qualitative criteria) the charge according to which the relation between rigid designator and designatum comes by magic. But rigidity and its attending transworld identity do run into genuine conflict with certain metaphysical positions about what we're talking about when we engage in counterfactual talk. The most prominent alternative to the commitment construes modal discourse about you in terms ofcounterparts of you who are not you and who act in your stead at other possible worlds. Thus, for David Lewis and fellow counterpart theorists, what matters to whether you might have driven instead of walked is indeed whether, in other possible worlds, someone other than you—but sufficiently similar qualitatively to you—drives instead of walks. So for Lewis, you don't have an important role, or indeed any role, among the players of the driving-as-opposed-to-walking worlds: and your qualities do. You are altogether absent from any world we entertain in considering the driving scenario as a scenario that you might have opted for, and in considering such a scenario we are not entertaining a thought about you as the protagonist. Let us call this proposal concerning counterfactual talk “serious counterpart theory” (following Russell 2013, p. 87).
The proposal that it is not you yourself but similar “counterparts” to you, who are relevant in this way to what you yourself might have done, is unappealing at face value: intuitively, as Noonan says, a claim about what you opted not to do but certainly “might have done hardly seems, at first sight, to be correctly interpretable as a claim about what someone else (however similar…) does in another possible world” (§6 of the entry on identity: this is a cleaned-up version of the so-called “Humphrey objection”: Kripke 1980, p. 45n.13; 1971, p. 148; cf. Lewis' replies to the Humphrey objection as originally cast: Lewis 1986, p. 196).
Indeed, one might go further by way of objections to the counterpart-construal of modal discourse, as Plantinga and Salmon do. Plantinga argues at length that the counterpart theorist's mistake is not really to misinterpret ‘you might have driven’, by offering an erroneous analysis of what makes it true; rather, the counterpart theorist's mistake is two-fold: the first mistake is toreject what ordinary speakers are saying by ‘you might have driven’, and the second mistake (a deviation from prudence) is to conceal the disagreement. The concealment is effected by assigning to the very same sentence used by ordinary speakers, ‘you might have driven’, an alien interpretation that the (serious) counterpart theorist can endorse, rather than the interpretation that would in fact be faithful to what the ordinary speakers are really saying. The new counterpart-involving interpretation renders reinterpreted sentences, no longer English, agreeable or offensive to the metaphysical commitments of a counterpart theorist pretty much when the respective English sentence does likewise for English speakers. In this way, the counterpart theorist can affirm and reject the usual modal sentences of English, understanding them differently though: so the counterpart theorist maintains the deceptive appearance of agreement with ordinary speakers by means of “a verbal camouflage,” as Salmon agrees (Salmon 1981, p. 236; cf. Stalnaker 2003, pp. 118ff.).
For her part, the counterparts-invoking theorist might emphasize the desirability of reduction.[6]Or she might point to other desiderada: for another likely example, she might wish to maintain contingency where rigidity commits us to necessity (Gibbard 1975, p. 188; Lewis 1986, p. 256; Delia Graff Fara joins many others in counting this as “the main benefit” of a counterpart construal: 2008, p. 186). One might turn to counterpart theory for relief from vagueness, since whether an object essentially has any given property can be a vague question; on the other hand, so too can the “question whether an object actually has a certain property,”as Kripke suggests (1980, p.115n.57). There are other motives.
The issues at stake between serious counterpart theory and transworld identity become complex quickly. The tangle of considerations can be more formally motivated than are the intuitive considerations already canvassed. Your theory might force unwelcome baroqueness in expression, when formalized (Linsky and Zalta 1996; Williamson 2000; 2002). Worse, you might find that you cannot even say in the preferred formal language what you want, intuitively, to say: e.g., “There could have been things that do not actually exist” (Russell discusses the history of this example: 2013, p. 485).
If its defenders are right, then counterpart theory can express all that anyone wants to express—even what transworld identity theorists want to express (Bacon forthcoming; Russell 2013). If so, then such qualms about formal resources need not divide serious-counterpart theorists from transworld identity theorists (and similarly for transtime identity theorists like Quinn 1978 (Meyer 2013)). But that would not reduce any divisiveness in the foregoing intuitive metaphysical considerations (these are the considerations that “must always play a dominant role,” too, according to Kripkean methodology anyway: 1976, p. 411).
الرجوع الى أعلى الصفحة اذهب الى الأسفل
مُشاطرة هذه المقالة على: reddit

Rigidity for Temporal and other Non-Alethic Modal Talk :: تعاليق

لا يوجد حالياً أي تعليق
 

Rigidity for Temporal and other Non-Alethic Modal Talk

الرجوع الى أعلى الصفحة 

صفحة 1 من اصل 1

 مواضيع مماثلة

-
» RIGIDITY
» The Modal Argument
» Philosophical Work for Rigidity.
» Relations Between Rigidity and Associated Theories of Reference
» 2.4 Non-modal Syllogistic

صلاحيات هذا المنتدى:تستطيع الرد على المواضيع في هذا المنتدى
** متابعات ثقافية متميزة ** Blogs al ssadh :: Pensée-
إرسال موضوع جديد   إرسال مساهمة في موضوعانتقل الى: