Bringing transparency to the de se debates

ABSTRACT

There has been much discussion over the last decade between so-called ‘de se skeptics’ and ‘de se exceptionalists’. The debate concerns claims made by John Perry and David Lewis in 1979 on the basis of some now-famous examples they introduced. In this paper, I argue that different problems and questions have commonly been conflated when approaching these issues, and I reframe the whole debate by appealing to two theses on the transparency of mental content that have been widely overlooked. These theses are essential components of the view challenged by Perry and Lewis and are indispensable for understanding the proper nature of the debates and positions in the literature. After distinguishing among three frequently conflated questions, I conclude that Perry’s and Lewis’s examples indeed present a problem for the view they were targeting, but this is not necessarily related to self-locating thoughts or indexical expressions. I also argue that our commitments to potential ‘de se thoughts’ are contingent on our positions regarding the transparency theses. I conclude the paper by exploring the prospects of an approach that, unlike most proposed views, completely abandons the idea that thoughts are relations between agents and contents.

KEYWORDS:

• Essential indexicals
• de se thoughts
• transparency of content
• propositions
• thought individuation

Acknowledgements

I presented parts of this article at the Open Philosophy Seminar and the ILCLI Seminar in Donostia, as well as at the ECAP11 Congress in Vienna; I would like to thank all the organisers and participants. I would also like to thank two anonymous referees for their kind comments and suggestions. Andrea Raimondi, Beñat Esnaola, Chris Genovesi, and Daniel Skibra read previous versions of the paper and made much-appreciated contributions to it. Special thanks are due to Miguel Ángel Sebastián; the ideas for this article originated in discussions I had with him about alleged essential indexicals and perspectival thoughts, and his criticisms and suggestions at different stages of the paper shaped the final result that I present here. This work was supported by the Department of Education, Universities and Research of the Basque Government under Grant IT-1612-22 and the Spanish Ministry of Science and Innovation, under Grant PID2019-106078GB-I00.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 I borrow the ‘de se skeptics/exceptionalists’ labels from Ninan (2016). On the skeptic side, we have Spencer (2007), Cappelen and Dever (2013), Devitt (2013), and Magidor (2015), on the exceptionalist side, Ninan (2016), Torre (2018), Shaw (2019), and Gray (2022).

2 See Perry (2019) and Falk’s (2015) review of Cappelen and Dever (2013) for similar opinions.

3 I distinguish between ‘self-locating thoughts’ and ‘de se thoughts’. By ‘thought’, I mean so-called ‘propositional attitudes’, like beliefs, desires, or judgments. Self-locating thoughts are those typically communicated by uttering sentences containing indexical expressions. My belief that I am working on a paper right now, or my desire for the sun to shine today are self-locating thoughts in this sense – I do not think anybody denies there are such thoughts. De se thoughts are thoughts that display a characteristic and differentiated nature in comparison to other non-de se thoughts – I say more on such de se thoughts in section 3. It is a substantial debate whether self-locating thoughts are, in this sense, de se.

4 It is debatable whether Perry and Lewis had the same target and, more particularly, whether the Doctrine, as characterised in this section, was the target of Lewis’s arguments. They both argue against the view according to which thoughts are relations to propositions, even though they do not share a common understanding of propositions. Perry states that, according to the Doctrine, propositions are ‘fine-grained’. Lewis, on the other hand, takes propositions to be sets of worlds, and these are famously ‘coarse-grained’. Magidor (2015) argues that it is such understanding of propositions which is on the origin of the problems introduced by Lewis, but, anyway, Lewis himself thought that he was introducing a new, different problem: 'You may think it goes without saying that the objects of attitudes are not sets of worlds because believing that 2+2 = 4 is not the same as believing that 123+456 = 579 though both equations hold at exactly the same worlds – namely, all. (…) My hunch is that this problem cuts across the issues I want to discuss, so I shall ignore it' (Lewis 1979, 515). Perry and Lewis understood that their examples posed a problem for the view that considers thoughts to be relations to propositions, and the problem arises even though we assume a fine-grained understanding of propositions (see section 2 of this paper). Furthermore, Lewis thought that the coarse-grainedness of propositions ‘cuts across’ the problem he was introducing. I will therefore assume that they were introducing the same problem for the same view.

5 The tenets identified by Perry naturally follow from this. Given that a thought is either satisfied or unsatisfied depending on how the world turns out to be, it is natural to conclude that the proposition reflecting its intentional character has a truth-value in an absolute sense. Tenet 3 states that propositions are fine-grained, even though it does not state how fine they have to be. Perry’s examples suggest that they should be fine enough to reflect the cognitive roles of thoughts.

6 A note on terminology: I distinguish between contents and propositions. As I understand them, contents are theoretical notions we stipulate, first and mostly, to account for intentionality – i.e. an object, state, or event represents the world insofar as it has content. Propositions are a sort of content, those of which it makes sense to state that they are true or false. Since I consider content to be the broader notion (propositions are contents, not every content is a proposition), from now on I will primarily use ‘content’ instead of ‘proposition’. Nevertheless, when discussing alternative proposals, I may employ other terminology (as in this section when discussing the Doctrine of Propositions).

7 ‘Transparency’ has been used in the philosophical literature to mean different things; perhaps its most famous use is due to Evans (1982), when discussing self-knowledge – here, I do not mean that. My use of the term is closer to accounts that claim meaning or linguistic content to be transparent:

It is an undeniable feature of the notion of meaning – obscure as that notion is – that meaning is transparent in the sense that, if someone attaches a meaning to each of two words, he must know whether these meanings are the same (Dummett 1978, 131).

Such principles on the transparency of meaning relate knowledge of meaning with the discriminatory capacities of speakers. According to them, a speaker can know, on the sole basis of their linguistic competence, whether two different expressions have the same meaning. Hence, this general idea is often presented in the form of two different claims, one concerning knowledge of sameness of meaning and another concerning knowledge of difference of meaning. Dummett’s quote concerns meaning or linguistic content, but some have proposed similar theses concerning mental content, especially in discussions on semantic externalism and self-knowledge (see, for example, Boghossian [1994]; Brown [2004] and ch. 6 in Stalnaker [2008]). Here are, for example, two such theses:

Transparency of sameness of content: for any two thoughts, or thought constituents, that S entertains at time t, if they have the same content then, at t, S can realize a priori that they have the same content.

Transparency of difference of content: for any two thoughts, or thought constituents, that S entertains at time t, if they have different contents then, at t, S can realize a priori that they have different contents (Brown 2004, 160).

I find such versions of the theses of the transparency of mental content problematic, for reasons that I cannot develop here. Let me just mention that such versions of transparency seem to me to implicitly assume an observational model of self-knowledge – they seem to assume that thoughts and their contents are things or events that we can observe in the 'theatre of our mind'. But, unlike linguistic expressions, thoughts are not entities that we somehow encounter somewhere; they are the states through which we cognise the world we live in. I fail to see how we can make discriminatory judgements about them. That is why I formulate the principles of transparency the way I do. We discriminate between things and events by means of our behaviour (by means of our linguistic behaviour, when we make explicit linguistic judgements): if we think that x = y, we will be disposed to show the same behaviour towards x and y, and if we think that x ≠ y, we will be disposed to show, at least in some conditions, different kinds of behaviour towards x and y. Hence, in the version of the transparency theses that I offer in the text, I exploit the idea that we may ‘discriminate by behaviour’. It may be argued (it has been argued to me in conversation) that the two theses I introduce are not strictly transparency theses. Well, I have tried to explain in this long endnote why I use such a label, but I take the discussion to be mainly terminological – I will not delve into this discussion, then.

8 As argued by Perry (2019, ch. 3), the issue has nothing to do with opacity, but cognitive significance. I will ignore this here.

9 There is a thread I will not pursue here. Although the ‘refining recipe’ has its own problems (see, for example, Kripke [1979]), the general idea is fairly clear for typical Frege cases: Lois’s two thoughts have two different general propositions as content, each proposition being in part constituted by different properties or senses satisfied by CK. But what general proposition can be offered as a candidate for the content of my mess if it has to be different to the content of any non-self-locating thought? This is an important difference between Perry's examples and their 'Frege counterparts', but, again, I will not pursue this thread here.

10 See Perry (2019, 16–19, 23-26, and ch. 8).

11 Spencer (2007) and Ninan (2016) also draw a parallel between the Heimson/Hume example and Twin Earth examples, arriving at similar conclusions. While Perry (2019) does not mention Twin Earths, he states that similar points can be made using ‘nambiguity’, where different objects have the same name.

12 Obviously, answering negatively to Q2 does not imply that indexicals or self-locating thoughts do not pose any interesting philosophical issue. Perry (2019), for example, states that an important part of the issue was explaining ‘why indexicals can disclose something important about beliefs not captured by singular propositions’ (17), how the meaning of indexicals relates to the cognitive significance of sentences containing them. I do not intend to discuss that here.

13 A fourth policy is logically possible: abandoning Sameness while keeping Difference. I know of no position that would fit under this description, nor can I think of any reason to adopt it.

14 See, for example, Burge (1977; 1993), Evans (1981; 1982), and McDowell (1986).

15 Torre (2018) does not discuss the view of a classical Neo-Russellian like Salmon (1986) but Perry’s (1979) own. But I am not certain about how we should interpret Perry’s (1979) proposal, either as viewing thoughts as ternary relations between agents, objects of belief, and states of belief or as rejecting that thoughts can be factored as relations to objects of belief. Therefore, I stick here to a classical Neo-Russellian like Salmon (1986).

16 For a ‘classificatory’ and ‘multi-level’ approach to satisfaction-conditions of mental states, see Perry (2001, 20–24, 92-99, 156-161), Korta and Perry (2011, 158–160), and de Ponte, Korta, and Perry (2023).

17 See Nelson (2019).

Additional information

Funding

This work was supported by the Department of Education, Universities and Research of the Basque Government [grant number IT-1612-22] and the Spanish Ministry of Science and Innovation [grant number PID2019-106078GB-I00].

Notes on contributors

Ekain Garmendia-Mujika

Ekain Garmendia-Mujika is an Assistant Professor in the Department of Philosophy at the University of the Basque Country (UPV-EHU) in Donostia-San Sebastian. He is also a member of the ILCLI Institute for Logic, Cognition, Language, and Information, and the Language and Thought (LAT) research group. He mainly works on the Philosophies of Mind and Language, on issues related to the individuation of mental content.