Abstract
It has become popular of late to identify the phenomenon of thinking a singular (or de re) thought with that of thinking with a mental file. Proponents of the mental files conception of singular thought (the MFC) claim that one thinks a singular thought about an object o iff one employs a mental file to think about o. I argue that this is false by arguing that there are what I call descriptive mental files, so some file-based thought is not singular thought. Descriptive mental files are mental files for which descriptive information plays four roles: determines which object the file is about, if any, it sets limits on possible mistakes that fall within the scope of successful reference for the file, it acts as a ‘gatekeeper’ for the file, and it determines persistence conditions for the file. Contrary to popular assumption, a description playing these roles is consistent with the file-theoretic framework. Recognising this allows us to distinguish the notion of singular thought from that of file-thinking and better understand the nature and role of both.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
Recanati (2012), 34, writes, ‘A non-descriptive mode of presentation, I claim, is nothing but a mental-file’.
This definition entails that singular thoughts are object-dependent. The definition has lost popularity in the recent literature in part for this reason, since it has become common to dispute this claim. See, for example, [Sainsbury (2005)].
Let’s assume that the italicized name here plays the role of an individual constant in the formalisation of this content. Its semantic role is to introduce an object for predication.
Or indeed like: (∀y) [(American President (y) → born in America (y)], which is also general. (1) can be thought of as a limiting case of general content.
One can give this definition but remain neutral as to whether the metaphysics of singular content is such that it contains an object itself as a constituent, or an object-dependent concept. In other words, both Russellians and Fregeans about singular content can adopt this definition.
The extent to which Jeshion rejects the definition of a singular thought as a mental states with singular content is unclear, even though this definition sits uneasily with her account. She is committed to denying that singular thoughts are object-dependent insofar as she allows that empty file-based thoughts are singular (see for example her ‘Vulcan’ example in [Jeshion (2002) & (2010)]). On the other hand, she does tentatively endorse the claim that singular thoughts are mental states with singular content, along with an object-dependent conception of singular content in [Jeshion (2010), 108]: ‘The content of singular thoughts are singular propositions, containing as their constituents individuals and properties.’ Therefore, although she explicitly but tentatively accepts the definition of singular thought as thoughts with singular content, the definition is inconsistent with her view.
A similar point about the limitations of defining singular thoughts in terms of their content is also made by Jeshion (2010). Her point here is that agreeing on the idea that singular thoughts have singular content does not settle questions about what it takes, cognitively, epistemically, etc., to think such thoughts.
The suggestion that we can use our everyday practices of ascription to settle this debate is mistaken because ordinary purposes attitude ascriptions do not systematically track the content of the attitudes they are use to ascribe. In Goodman (unpublished. a), I argue that the truth conditions of ordinary-purposes attitude ascriptions underdetermines even the truth-conditional content of the mental states they are used to ascribe. For other discussions of the way that ordinary ascriptions fail to track the content of the mental states they are used to attribute see [Recanati (2012) 150–54], [Bach (1987), (1997)] & [Taylor (2002)].
Jeshion (2010), 108, agrees: ‘For descriptive thought there is widespread agreement [about what it takes to entertain such thoughts]: One must possess and grasp those constituent concepts in the general proposition, must do so in the way in which they are structured in the proposition.’ See also, [Jeshion (2010), 129].
This terminology is derived from Bach’s distinction between thoughts whose objects are determined satisfactionally vs. relationally (see [Bach (1987))]. Bach is also an MFC proponent.
This fits with Recanati’s claim that files ‘are a matter of information clustering’ ([Recanati (2012), 42]).
Again, this is in line with Recanati’s recent conception of files according to which the existence of a file licenses the integration and exploitation in automatic inference patterns of information stored in that file (See, for e.g., [Recanati (2012), 41, 96]).
It is important that, in spelling out this example, I intend no commitment to the idea that (3) & (4) state the contents of beliefs or thoughts. (3) and (4) are rather statements that might be used to express the beliefs or thoughts involved. According to proponents of the mental-files framework, the fact that the thought one has at the office about the individual whose office is across the hall and the thought one has at home about the individual who lives next door both involve the use of the same mental file (as well as predication of the same property). On the mental-files framework, this means that these two thoughts have the same content. As I explain further below, this is because, according to file-theorists, a mental file is an object-concept, and the two thoughts employ the same object-concept. Thanks to an anonymous reviewer for requesting that this point be clarified.
For more discussion of what this means, see Goodsell (2013). The sense of ‘trade on identity’ I use here is (roughly) like Goodsell’s ‘current-reasoning reading’, according to which one trades on the identity of a and b if one is disposed to reason as if a = b without using an additional identity premise in one’s reasoning. I don’t mean to suggest, by appealing to this role for mental-files, that we have at our disposal a fully worked-out sense of what it means to reason without an extra premise, or an adequate understanding of when this occurs. This is work that must be done in order to fill out the theory of mental-files. What does seem clear, however, is that there is that reasoning of this kind must occur for, without it, it is unclear how justified inference would be possible (see [Campbell (1987]).
See [Perry (2000), 69]. It is important to note the following: Although in some of Perry’s examples there is no continuing descriptive information available to the subject, which could be used to specify the putative (continued) belief, this does not entail that it is only in cases in which there is no continuing descriptive information that files can be used to theorise continuing belief. Perry’s claim is that continuation of either descriptive information or (in his terms) ‘text’ is not a necessary or sufficient condition on continued belief (Ibid., 81). This is consistent with the claim that continuity of a file accounts for continuity of belief in cases where there is continuing descriptive information available, or a continuity of ‘text’. This point is often overlooked, and is in fact related to my later claim that it is a mistake to think that the file framework is only suited to theorising cases of non-descriptive thought.
This is sometimes put by saying that files play the role of modes of presentation (See, [Recanati (2012), Ch. 3])
This is somewhat complicated by the possibility (contra Strawson) of knowingly keeping twodistinct but co-referential files. There are several reasons why this might occur (without irrationality): for example, relatively low credence in the identity fact (which nonetheless rises to the level required by knowledge or belief), or it being useful to keep track of the perspectives of others (or one’s former self) on the object [Recanati (2012), 44-45]. A natural way to theorise such cases is in terms of files that are linked, but not merged. There are several interesting issues attached to the possibility of such cases, but I set them aside here.
That my file on an object o loses or gains a predicate ‘is F’ from t 1 to t 2 does not preclude that the file accounts for my continuing to believe, from t 1 to t 2 , that o is G. However, in cases of complete overhaul of descriptive information stored in a file, that file’s accounting for continued belief will be foreclosed. It is a constraint on a file’s being used to theorise continuation of the belief that o is G that ‘is G’ remains in the file for the duration of the belief (otherwise there is no continued commitment to o being G). In this sense, ‘Ship of Theseus’ cases are marginal, in that they are cases in which a file cannot play one of its theoretical roles. In such cases, the file can account for trading on identity in inference, and continuation of object-concept, but not for continued belief.
It is arguable, but not obviously true, that a file could be a file on a even if it contained no predicates true of a.
A note on two absences from this list: Firstly, the connection between files and name-use does not register as an item on the list of the features of files. This is for a reason discussed in the final paragraph of §II. Even those who emphasize the connection between files and name-use do not argue that all files are associated with names. Secondly, the reader may wonder whether the idea that files involve rigid reference to objects ought to be included as a feature of mental-files (Thanks to Heather Logue for raising this possibility). This, one might suggest, is another reason to think that file-based thoughts are singular. My position, however, is that files do not always refer rigidly to their objects. Although my argument in this paper does not focus on the notion of rigidity, I argue this in [Goodman (unpublished. b)] where I also give a case of a non-rigid file.
The particular notion of governing I will spell out here is inspired by points stressed by Dickie (2011), in giving a theory of how proper names refer, although my conclusions are not the same as hers. In Dickie (2011), a distinction is made between cases in which mistaken descriptive information (in particular, mistaken kind, or sortal, information) interfere with the referential properties of a name, and cases in which such mistakes do not. My notion of governing spells out a relation that a piece of information can bear to a file containing it, such that it does determine the referential features of the file along with other central features of the file (to be spelled out in what follows). Dickie’s cases in which a piece of descriptive information is mistaken but this does not interfere with the referential properties of the containing file are cases in which that information does not govern the file. A note on terminology: Despite the fact that my notion of governing is inspired by Dickie’s cases, it is very different from, and therefore not to be confused with, the notion of governance used by Dickie herself (see, again [Dickie (2011)]). My notion of governing spells out a (privileged) relation that a piece of descriptive information can bear to a file in which it is stored. Dickie’s notion of governance is a relation that an object can bear to a file, or body of beliefs.
When I use the term ‘reference’ in relation to mental-files, I do not use it in the ‘strong’ sense (sometimes contrasted with denotation) that implies singular reference.
I’m going to largely set aside questions about the status of thoughts that employ empty files for the purposes of this initial exploration of descriptive files. However, we can work with the assumption that these thoughts are like other descriptive thoughts: in the case where there is no unique F, descriptive thoughts that employ a file governed by the description ‘the F’ are false.
There is disagreement among MFC proponents about the constraints on descriptive reference fixing for files. Jeshion (2010) allows for descriptive reference fixing without acquaintance as long as her significance condition is fulfilled. Recanati (2012) has a more complex position. He thinks that acquaintance is a normative requirement on opening a file but, simplifying somewhat, allows for descriptive reference-fixing on the condition of either expected or ‘imagined’ acquaintance (see [Recanati (2012), 167 & 168], respectively).
I take this helpful terminology from Dickie (2011) & (MS).
The perceptual connection associated with a perception-based file is an instance of what Recanati (2012) calls an ‘epistemically rewarding relation’.
That is, abstracting away from the predicative part of their content: the semantic content of a thought employing a descriptive file governed by the description ‘the F’ is a descriptive content: ‘The F is ....’
Again, I abstract away from the predicative part of the content here.
I am allowing myself to talk here as if the information cluster can be distinguished from the file in order to coherently describe the envisioned scenario. To the extent that this causes discomfort or tension, this suggests that the idea of a descriptive file entails counterintuitive persistence conditions for the files it is applied to. This challenge is discussed later in the paper, in section §VI. However, it should also be noted, firstly, that the diachronic identity conditions for files are a difficult issue requiring separate treatment and, secondly, that there are arguments that a combination of continued information cluster with difference of file should be allowed and can be accounted for within the file-theoretic framework in terms of a so-called ‘conversion' operations by which information from one file is transferred into another (see Recanati (2012) ch.7 for discussion).
Contrast the case of a perceptual, demonstrative file, where the criterion for which information is stored in the file is that it must be information accessed through the perceptual connection associated with the file.
Recanati (2012) seems to imply this kind of view at times (e.g., see 34–35 & 165)
Some descriptive files may be opened and maintained to keep track of the satisfier of a definite description, without the introduction of an accompanying (descriptive) name.
Let me note in advance that, by claiming that some mental-files associated with descriptive names are descriptive files, I do not thereby claim that descriptive names (the linguistic devices belonging to a public language) are not singular referential linguistic terms. I also need not claim that all files associated with descriptive names are descriptive, rather than singular.
In this paper, I am not concerned with all the details of Jeshion’s own account of singular thought, but rather with the broader question of whether file-based thoughts are always singular. In [Goodman (unpublished. b)] I give an argument against Jeshion’s particular version of the MFC.
Jeshion has her own stock of favorite examples of file-based thoughts involving descriptive reference fixing [Jeshion (2010), 116–117]. I choose not to use one of these particular cases because they all introduce complicating factors such as the central role of causal traces of the putative referent of the file (e.g. Unabomber, Bearprint), emptiness (e.g. Vulcan), or future directed reference (e.g. Dessert Sensations). I choose to focus instead on a simpler case to elucidate the notion of a descriptive file. However, but will return to discuss one Jeshion’s own cases for later in the paper.
If one were committed to the claim that a mental-file on an object can only be formed when one has a causal, informational connection to that object, one might react by denying that this could be a case of file-thinking, but I do not find this denial plausible. To the extent that a mental-file is a matter of information clustering, it is unclear why a case of information clustering could not begin with an act of descriptive reference fixing, rather than acquaintance. Even proponents of the MFC who hold that mental-files require acquaintance (e.g., [Recanati (2010) & (2012)]) also attempt to accommodate the possibility of this kind of case.
Jeshion (2009) (2010) has disputed both that (Evans’s) ‘Julius’ is a genuine descriptive name, and that Evans’s case would meet the conditions for opening a mental file on the inventor of the zip. She argues that there is a significance condition on both naming and singular thought, which is not satisfied in Evans’s (1982) description of the case. Here, I amend the case such that Jeshion’s significance condition is satisfied.
Although de Mestral did invent Velcro, the rest of this story is fictional embellishment.
If one is concerned that the descriptive nature of the case is undermined by de Mestral’s contact with causal traces of the zipper’s inventor in the form of perception of instances of his invention, we could stipulate that de Mestral has never seen a zipper.
We can grant that, without the intention or possibility of forming beliefs about the object in question, and in the absence of any particular psychological dispositions with respect to reasoning about that object, this would not be a case of forming and maintaining a mental-file.
It is a priori for him that, if his rival exists, he is the inventor of the zip.
The fact that de Mestral’s conception of the inventor of the zip is somewhat fanciful and distorted adds to the intuition that a fairly thin description, and a peculiar sense of what follows from this, plays its gatekeeper role. As we’ve described the case, de Mestral has not made any real effort to consider or learn about his rival’s identity beyond his status as the inventor of the zip.
This is not a totally straightforward claim, since there may be sortal or categorial restrictions on referential success, even for the case of many singular files. This might entail that even singular files can’t contain entirely false information. This would depend on the precise nature of the sortal or categorical restrictions: it is only entailed if the form of the sortal restriction was not just that certain sortal errors could not occur, but rather that certain sortal information was required to achieve reference. For an argument against the view that this is a requirement on all singular thought, however, see [Goodman (2012)].
Again, we abstract here and in the rest of this paragraph from the predicative part of the content: Thoughts employing a descriptive mental file governed by ‘the F’ will have the content, ‘the F is …’
Thanks go to Simon Prosser for helpful discussion of this question.
Jeshion’s claim is not that the mere fact of an act of descriptive reference-fixing entails that a file is opened and maintained. She claims that ‘significance’ is a condition on the possession of a (non-perceptual) mental-file (See [Jeshion (2010), 136]).
It is a priori for him that, if Dessert Sensations exists, it is a cake delivery business.
In particular, counterintuitive persistence conditions can be dealt with via a story similar to Recanati’s (2012) story about conversion, which is an operation on files by which information from a file that ends is placed in a newly created file. A story about conversion allows us to accommodate intuitions about continuation of an information-clustering pattern without appealing to continuation of file. Although I lack space to fully address the merits of this strategy here, there may be good reasons to think we will need to appeal to it in accounting for the diachronic identity conditions of files, even for cases of singular files (I explore this issue in work that is currently in progress).
As I’ve noted, it isn’t straightforwardly the case that this is possible for perception-based files, but let us assume for purposes of argument that it is. See n.45 of this paper, and also the considerations raised in [Dickie (2011)].
References
Bach, K. 1987. Thought and reference. Oxford: Oxford University Press.
Bach, K. 1997. Do belief reports report beliefs? Pacific Philosophical Quarterly 78: 215–241.
Campbell, J. 1987. Is sense transparent? Proceedings of the Aristotelian Society 88: 273–292.
Dickie, I. 2011. How proper names refer. Proceedings of the Aristotelian Society 111: 43–78.
Evans, G. (1982) The Varieties of Reference. J. McDowell (ed.), Oxford: Oxford University Press.
Evans, G. 1985. The causal theory of names, reprinted in Gareth Evans: Collected Papers. Oxford: Oxford University Press.
Goodman, R. 2012.‘Why, and How, Not to be a Sortalist about Thought’ Philosophical Perspectives Volume 26
Goodsell. 2013 ‘Mental files and their identity conditions’, Disputatio, Book Symposium on Recanati’s Mental Files V, 36, 177–191.
Grice, H. (1969.‘Vacuous names’ in D. D. J. Hintikka (ed.) Words and Objections, 118–145. Reidel.
Heim, I. 1982. The Semantics of Definite and Indefinite Noun Phrases. Ph.D. thesis, University of Massachusetts, Amherst.
Jeshion, R. 2002. Acquaintanceless de re Belief. In Meaning and Truth: Investigations in Philosophical Semantics, ed. J. Campbell, M. O’Rourke, and D. Shier. New York: Seven Bridges Press.
Jeshion, R. 2009. The significance of names. Mind & Language 24: 372–405.
Jeshion, R. 2010. Singular thought: acquaintance, semantic instrumentalism, and cognitivism. In New Essays on Singular Thought, ed. R. Jeshion. Oxford: Oxford University Press.
Kripke, S. 1980. Naming and Necessity. Cambridge: Harvard University Press.
Lockwood, M. 1971. Identity and reference. In Identity and Individuation, ed. M. Munitz, 199–211. New York: New York University Press.
Perry, J. 2000.‘A Problem About Continued Belief’, in The Problem of the Essential Indexical and Other Essays. Stanford: CSLI.
Recanati, F. 2012. Mental files. Oxford: Oxford University Press.
Sainsbury, R.M. 2005. Reference without Referents. Oxford: Oxford University Press.
Taylor, K. 2002. De Re and De Dicto: Against the conventional wisdom. Noûs 36: 225–226.
Acknowledgments
Thanks to Imogen Dickie, Aidan Gray, Simon Prosser and Robbie Williams for discussion of this and related material. Thanks also to Jeff King for comments on an earlier version of the paper, to an audience at the Leeds Philosophy Department Senior Seminar for their questions, and to two anonymous referees at Review of Philosophy and Pscyhology for helpful comments.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Goodman, R. Against the Mental Files Conception of Singular Thought. Rev.Phil.Psych. 7, 437–461 (2016). https://doi.org/10.1007/s13164-014-0209-0
Published:
Issue Date:
DOI: https://doi.org/10.1007/s13164-014-0209-0