¹ Evans takes himself to be offering an unified account of demonstrative thoughts, which to him comprise ‘this’-thoughts, ‘here’-thoughts, as well as ‘I’-thoughts. My concern in this paper is exclusively with his theory as a theory of first-person thought (and ‘here’-thought to some extent).

² Evans, Gareth, The Varieties of Reference (Oxford: Clarendon Press 1982)Google Scholar. Hereafter VR.

³ Evans, VR, 183.

⁴ Evans, VR, 207.

⁵ The 90's are an era that saw the development and fine-honing of a type of argument against perceptual models of self-knowledge. Sydney Shoemaker has done more than anyone else in arguing against the perceptual model. See Shoemaker, Sydney, ‘On Knowing One's Own Mind’ Philosophical Perspectives 2 (1988) Epistemology, 183–209 CrossRefGoogle Scholar; Shoemaker, , ‘Self-Knowledge and Inner Sense’, Philosophy and Phenomenological Research LIV (1994), 249–314 CrossRefGoogle Scholar.

⁶ The popularity of the EAIC reading is evident to anyone familiar with the literature on Evans. I'll be focusing on three particular writers below, but I could have easily singled out others who have written on Evans. See for example the collected essays on Evans, in Bermúdez, J. (ed.) Thought, Reference, and Experience: Themes from the Philosophy of Gareth Evans (Oxford: Clarendon Press, 2005)Google Scholar, see in particular Bermúdez's introduction to the essays, where he characterizes Evans theory of demonstrative thought (including ‘I’-thought) in terms of ‘ways of thinking about object’(4) ‘discriminative knowledge'(6) and ‘cognitive ability’(7). If what I say in this paper is correct, these terms will only give us half of what is required to characterize Evans’ theory of ‘I’-thought.

⁷ O'Brien, Lucy, ‘Evans on Self-Identification’ Nous 29:2 (1995), 232–247 CrossRefGoogle Scholar.

⁸ O'Brien, ‘Evans on Self-Identification’, 233.

⁹ O'Brien, ‘Evans on Self-Identification’, 236–37.

¹⁰ Stanley, Jason, ‘Names and Rigid Designation’, endnote 27, in Oxford Companion to the Philosophy of Language (Oxford: Oxford University Press 1997), 555–85Google Scholar.

¹¹ Stanley, ‘Names and Rigid Designation’, 581

¹² Hamilton, Andy, ‘Anscombian and Cartesian Scepticism’ The Philosophical Quarterly 41:162 (1991): 39–54 CrossRefGoogle Scholar.

¹³ Hamilton, ‘Anscombian and Cartesian Scepticism’, 54

¹⁴ Ibid.

¹⁵ Evans, VR, 207.

¹⁶ Ibid.

¹⁷ Evans, , ‘Molyneux's Question’, in Collected Papers (Oxford: Clarendon Press, 1985), 364–399 Google Scholar.

¹⁸ Evans, VR, 154.

¹⁹ Ibid., 390.

²⁰ John Campbell, ‘Information Processing, Phenomenal Consciousness, and Molyneux's Question’, in J. Bermúdez, op.cit., 195–219.

²¹ Both papers are collected in Evans, , Collected Papers (Oxford: Clarendon Press, 1985)Google Scholar.

²² Campbell has a more serious charge against Evans. He maintains that Evans conflates informational content with content of phenomenal experience. I can't go in the details here.

²³ Campbell, ‘Information Processing, Phenomenal Consciousness, and Molyneux's Question’, 200.

²⁴ Evans, VR, 158.

²⁵ Evans, VR, 158.

²⁶ Intention and intentional action (past or current) are clearly different notions, and I'm not suggesting, in giving this definition of practical knowledge, that they're not. My concern here is with the knowledge the agent has of her intention and intentional action, not with these notions per se, and as far as the nature of this knowledge goes, theories fall nicely under the two categories I wish to draw here. Hence my rather loose language in this regard in this part of the paper.

²⁷ Evans, VR, 210–12.

²⁸ Evans, VR, 207.

²⁹ Ibid.

³⁰ Here is the full text of the footnote (with reference omitted): ‘This element [viz, the action component] in an account of ‘I’ is stressed in much recent work: [Here Evans mentions two work by John Perry and one By D. Lewis]… Neglect, in this work, of the other element [viz, the information component] produces a strangely one sided effect – ‘strangely’, because the other element is just as striking, and clearly parallel, and also because the dominant conception of the identification of empirical content concentrates exclusively on the input or evidential side of things. This chapter will partly redress the balance by rather neglecting the action component.' (VR, 207). In hindsight, this has to be an unfortunate strategy, which has undoubtedly contributed to the perception that Evans operated with a perceptual model of ‘I’-thought. The citation of Perry among others as those who have addressed the action aspect of ‘I’-thought is also an oversight on Evans' part. The disagreement between Evans and Perry on the treatment of demonstrative thought in general is well-known, but by referencing Perry as one of those who have dealt with the action component, Evans also failed to appreciate the full extent of difference between him and Perry on the action issue, for unlike Evans, there is little in Perry's notion of action that remotely resembles Evans' conception of intentional action, to be laid out below.

³¹ By ‘causal theories’ I mean to include any and all theories which take actions as events with a cluster of causal ancestry factors, be these belief, desire, belief-desire complex, belief and desire caused intention, or even ‘the agent’ itself (the ‘agent’ in ‘agent causation’ theories for example).

³² Anscombe, E., ‘The First Person’, in Guttenplan, Samuel (ed.), Mind and Language (Oxford: Clarendon Press, 45–65), 87Google Scholar.

³³ McDowell, John, ‘Anscombe on Bodily Self-Knowledge’, in Essays on Anscombe's Intention (ed.) Ford, A., Hornsby, J., and Stoutland, F. (Harvard, Mass.: Harvard University Press, 2011)Google Scholar. McDowell attributes the idea to the work of Sebastian Rödl on Aristotle. Self-knowledge based on information received through the exercise of receptive faculties can at the best be knowledge of oneself as other, but not knowledge of oneself as self (or agent), and only the latter qualifies as a genuine form of self-consciousness.

³⁴ Anscombe, op.cit. It's true that Anscombe also allows certain non-practical knowledge to be non-observational (e.g. limb position), but her reason for this may be deeper than her words seem to convey. See McDowell, op.cit. for a subtle and sustained discussion of the matter.

³⁵ Evans, VR, 207.

³⁶ McDowell, , ‘Referring To Oneself’, in The Philosophy of P. F. Strawson, ed. Hahn, L. (Chicago and LaSalle, Illinois: Open Court Publishing 1998), 129–145 Google Scholar.

³⁷ McDowell, ‘Referring To Oneself’, 143.

³⁸ This is not to suggest that Evans already had McDowell's notion of agency as intention qua physical intervention. But given the intellectual closeness between the two thinkers, that there could be continuity of the sort should not be completely surprising.

³⁹ I can't go in the details here. The idea is basically this: behavioral interaction in the form of the exercise of a practical skill to keep track of an object in the same behavioral environment is what effects the switching of modal-specific information into a type of content general enough to be the content of thought. Continuity of content is a matter of having not lost track of the object one came into informational contact with at an earlier point. This is so with one's ‘this’-thought and ‘here’-thought, and it is also so with one type of ‘I’-thoughts – the type where the ‘I’ is based on receptive knowledge of oneself as an item in the objective order of things.

⁴⁰ These would be ‘I’-thoughts which are not immune to the two types of errors I mentioned in sec. 3.1. Evans was well aware of this aspect of what I here call ‘information-based ‘I'-thoughts' (see his discussion of the many types of abnormal circumstances concerning such thoughts in the later part of Chapter 7 in the VR). Wittgenstein's well-known notion of ‘use of ‘I' as object' is another way of putting the same idea here.

⁴¹ In her criticism of Evans, O'Brien made much of cases of information-based ‘I’-thoughts under abnormal conditions. Her sole reliance on such cases highlights the peril of ignoring the role of agency in Evans' or, for that matter, anyone's theory of first-person thought.

⁴² I would like to thank Fred Zammiello for reading an earlier draft of this paper.