Disagreement, Error, and an Alternative to Reference Magnetism

Bibliographic Data

ID5828964
AuthorsTimothy Sundell (University of Kentucky, corresponding author)
Year2011
Volume90
Issue4
Pages743-759
Publication date2011-09-16
Peer ReviewedYes
Open AccessNo
TypeARTICLE
VenueAustralasian Journal of Philosophy (JOURNAL)
Journal identifiersISSN: 0004-8402 • E-ISSN: 1470-6828
PublisherRoutledge (PUBLISHER • GB)
DOI10.1080/00048402.2011.614266
OpenAlexW2021905992
LanguageEN
Citations received15
References cited24

Lewisian reference magnetism about linguistic content determination [Lewis Citation1983 has been defended in recent work by Weatherson [2003] and Sider [2009], among others. Two advantages claimed for the view are its capacity to make sense of systematic error in speakers' use of their words, and its capacity to distinguish between verbal and substantive disagreements. Our understanding of both error and disagreement is linked to the role of usage and first order intuitions in semantics and in linguistic theory more generally. I argue, partially on the basis of these more general considerations, that reference magnetism delivers implausible results. Specifically, I argue that the proponent of reference magnetism maintains her analysis of genuinely systematic error at the cost of an empirically unjustifiable error theory regarding ordinary usage. In response, I describe an alternative view of content determination-MUMPS, or Meaning is Use Minus Pragmatics-which is not committed to such error theories. Despite this advantage, MUMPS has high prima facie costs. On such a view, there is a great deal of variation in linguistic meaning across speakers and times. As a result, a large number of seemingly mistaken claims are analysed as expressing true propositions. Correspondingly, a large number of seemingly substantive disagreements are analysed as terminological. However, I argue that these consequences are not as costly as they seem. Despite appearances, MUMPS is consistent with objective, metaphysically realist adjudication of disagreements, even in cases where meanings are not shared and where both parties to a dispute speak truly. MUMPS thus allows for a more nuanced understanding of linguistic usage, change, and variation, without imposing a commitment to any form of metaphysical anti-realism. Keywords: reference magnetismdisagreementtruthrealismsemanticspragmaticslinguistics Notes 1Hirsch Citation2009 contains one well-known example of this style of argument, concluding that metaphysicians with certain opposing ontological views use the existential quantifier in different ways. Sidelle Citation2007 considers the question about whales (among others), while Sundell Citation2010 takes up the aesthetic cases. Gibbard Citation1990 warns against certain analyses of rationality in the light of their consequences for disagreements about ethics, while the problem of moral disagreement more generally is the focus of its own lively and growing literature. In addition to these examples, many arguments for contextualist analyses in various philosophical domains are instances of the usage argument. Similar issues are at the centre of recent discussions of semantic relativism. 2Variations on the stability argument are applied across a range of philosophical subfields. See, for example, Hare Citation1952, Gibbard Citation1990, and Smith Citation1994 in metaethics, Lasersohn Citation2005 and MacFarlane Citation2007 in aesthetics, Manley Citation2009 in metaphysics, Richard Citation2004 in epistemology, and Cappelen and Hawthorne Citation2010 in philosophy of language, among many others. 3When I say relativism, I mean anti-realist relativism of the sort that reduces scientific debates and discoveries to terminological disputes and revisions. This use of relativism is to be distinguished from semantic relativism of the kind advocated in Lasersohn Citation2005, MacFarlane Citation2007, Stephenson Citation2007, and Egan Citation2007, among others. To avoid confusion, I will, with a few exceptions, use Lewis's phrase, over-easy truth, discussed below. 4Pullum Citation2005. 5See Wason and Reich Citation1979 and von Fintel Citation2004 for discussion. 6The proper semantic account of fiction and pretend is of course its own controversial issue, one that I won't engage here. For now I will be content simply to grant that in the clearest cases of pretend-when a child refers to a bicycle as a spaceship, for example-the linguist is free to disregard data that on the most superficial analysis would contradict an otherwise well supported theory. 7See Horn Citation2004 for discussion. This analysis of numbers is not uncontroversial, but serves well as an example. (See Chierchia Citation2004 for an alternative.) Alternative cases of quantity implicature are easy to come by. 8MUMPS is a view about semantic methodology and in particular about the facts against which the success of a semantic theory should be judged. It is not a view about the metaphysical status of meaning and I am not advocating a 'use theory' in the sense of Horwich Citation2004 or others in that tradition. A clearer label might be Meaning is (given by) Use Minus Pragmatics, though 'given by' is still infected with some of the epistemic/metaphysical ambiguity that plagues similar expressions like 'determine'. I hope that the mnemonic advantages of the label outweigh any potential confusion. Thanks to an anonymous referee for emphasizing this point. 9In his account of reference magnetism, Weatherson, like Lewis, is very much concerned to address problems about disagreement and over-easy truth. Historically, however, reference magnet views are motivated largely by the challenges of Kripkensteinian scepticism about meaning and Putnam's model theoretic argument. (See Lewis Citation1983, Citation1984.) Those challenges are related to the problem of over-easy truth, but they call for independent and specific responses. I do not have any such response to offer here. Thus, for the purposes of this work, I am content to argue for the partially conditional thesis: If MUMPS is wrong, it is not because it commits us to over-easy truth, because it doesn't. I believe that in fact there are responses to be made to the challenges of Kripke and Putnam on behalf of a MUMPS-type view, and I develop them in detail elsewhere. See Sundell [ms]. 10On Merrill's original suggestion, naturalness is an all-or-nothing issue. Lewis Citation1984: 227] jettisons that assumption and argues that naturalness, and thus reference eligibility, is a matter of degree. My arguments against reference magnetism will apply even given this more nuanced account of naturalness. Nevertheless, I am sceptical that the more nuanced account-for all its greater plausibility as a theory in metaphysics-can succeed as a theory of content determination, even setting aside the worries raised elsewhere in this paper. If naturalness comes in degrees, then what could it mean to say, as proponents of reference magnetism are wont to, that our use of a predicate is 'close enough to a natural category to fall into its orbit' or that our concept is 'in the neighbourhood of a natural category'? For any natural category C that our usage is close but not perfectly faithful to, there will be another category C*-only slightly less natural than C-that fits our usage more closely. How is the slightly more natural category meant to exert enough 'gravitational pull' to dislodge our meaning from the slightly less natural category that fits our usage slightly better? For now, I leave the issue aside. (Thanks to Andy Egan [p.c.] for discussion.) 11See, for example, National Oceanic and Atmospheric Association 2006 for a scientific document in which this fact and the corresponding usage is presupposed. The lake is referred to there, as in similar scientific documents, as Lake Michigan-Huron. Elsewhere it is sometimes called Lake Huron-Michigan. 12Or, in cases where the relevant proposition is presupposed, truth-valueless. 13This, emphatically, is not to deny that there is such a thing as metaphysical naturalness independent of our 'practices'. It is rather to observe that in some domains we do not appear to regulate the usage of our terms according to that standard. 14See, for example, Chomsky Citation1965. For more recent discussion, see Ludlow Citation2011. 15U.S. Department of Justice, Office of the Assistant Attorney General [2002]. 16To be clear, I do not ask the reader to assume that the speakers of (19) and (20) mean different things by torture. Rather, I argue that even on the supposition that they do, their dispute can nevertheless be analysed as substantive. 17What about the modern biologist who is also a sailor and who varies their usage accordingly? The same resources available to the proponent of reference magnetism are available to the MUMPSer in such cases, including ambiguity, polysemy, context-sensitivity and the like. 18An anonymous referee suggests that a dispute between a biologist and a sailor might nevertheless be substantive if there is disagreement as to which activity (or perspective) is the important or salient one. I think this is correct. However, if such a dispute is to make sense, we must imagine that there is some third, more general activity with respect to which this background dispute can be adjudicated. We might imagine, for example, that the dispute takes place in the context of writing marine environmental policy. In such a context, it would make sense for speakers to advocate for the relevance of the whaler's categories over the biologist's, or vice versa. If there is not a more general, controlling standard, however-if each speaker simply sticks to their own perspective because it is their own-then the MUMPSer will indeed analyse their dispute as defective. I submit that this is precisely the right result. 19Note that the difficulties raised by relative, rather than absolute, naturalness (discussed in note 11) are not a problem for MUMPS. Without the metaphorical talk of 'gravitational exertion' and 'neighbourhoods', the task of ranking concepts by respective metaphysical naturalness is unmysterious, if difficult in practice. 20Thanks to Clare Batty, Stefan Bird-Pollan, Josh Brown, Fabrizio Cariani, Ezra Cook, David Dick, David Ebrey, Andy Egan, Sam Epstein, Thony Gillies, Sandy Goldberg, Lee Goldsmith, Peter Ludlow, Alex Plakias, David Plunkett, Peter Railton, Meg Wallace, and two anonymous referees for helpful comments and suggestions

Adjudication · Philosophy of language · Prima facie · Variation (astronomy · Epistemology, Ethics, and Metaphysics · Philosophy and Theoretical Science · Syntax, Semantics, Linguistic Variation

Tools

Open DOISci-Hub