Précis: Perplexities of Consciousness


for Philosophical Studies




Eric Schwitzgebel

Department of Philosophy

University of California at Riverside

Riverside, CA  92521-0201


eschwitz at domain:



May 17, 2013




Précis: Perplexities of Consciousness


We do not know our own stream of conscious experience very well, nor is it easy to remedy our ignorance.  That is the thesis of Perplexities of Consciousness.

The method is to look, in some detail, at a wide variety of cases: the coloration of dreams, the visual experience of perspective, double vision, visual imagery, the auditory experience of echolocation, combination tones, afterimages, visual illusions, the tactile experience of your feet in your shoes, emotional experience, peripheral vision, cognitive phenomenology, and visual experience with your eyes closed.  About each type of case, I argue, we are prone to error or ignorance – in some cases, error or ignorance of a magnitude that would be shocking if we were trying to reach similarly large-grained judgments about nearby perceptible objects in the external world.  For example, we seem to miss, or alternatively invent, entire modalities of experience – cognitive phenomenology (a.k.a. imageless thought or unsymbolized thinking), or echolocation, or distinctively emotional phenomenology.  And we seem to miss, or alternatively invent, what one might have thought would be hugely obvious features of the experiences we do recognize, for example, that almost all of the objects that we see are visually doubled (or that they’re not), or that we have a wealth of sensory experience beyond the boundaries of attention (or that we don’t).  Imagine being mistaken about whether nearly all the ordinary-sized physical objects in your environment were single or double, or being mistaken about whether you were surrounded by a richly-populated abundant world or a thin, sparse, mostly empty world!

A long philosophical tradition, going back at least to Descartes, has held that we know our own minds, especially our current conscious experiences, better than we know the outside world – maybe even infallibly, or indubitably, or incorrigibly.  If the reasoning in my book is correct, that tradition is badly mistaken.  In the 20th century, the strictest forms of infallibilism fell out of favor, but most critiques were limited: Critics typically argued only against perfection, while allowing that self-knowledge might still be excellent, or they focused their critiques on self-knowledge of nonconscious aspects of our own minds, such as the causes of our decisions.  The most important exception to this pattern was the introspective psychology movement of the late 19th and early 20th centuries – especially E.B. Titchener (discussed at length in Chapter 5 of Perplexities) – but their arguments have been neglected.

I argue from specific cases to a generalization.  My arguments concerning specific cases tend to take one of three forms.

The argument from variability.  Group A says X about their stream of experience.  Group B says not-X.  It is not plausible that Group A and Group B have experiences different enough to make the claims of both groups true.  Therefore, at least one group must be mistaken about their experience.

The argument from error.  Group A says X about their stream of experience.  But X is probably false.  Therefore, Group A is probably mistaken about its experience.

Doubt induction.  I aim to induce personal doubt in the reader about whether her stream of experience has feature X, leading her to conclude she does not know whether her experience has feature X.  This doubt induction sometimes proceeds by asking the reader carefully to reflect on her own phenomenology, with some pointed questions to consider.

I hope for synergy among my arguments.  Suppose one type of argument works for one type of experience.  To the extent that argument finds success, it can lend support to other types of argument regarding that same kind of experience.  If I succeed in doubt induction about the reader’s own visual experience of perspective, for example, that might make it seem more plausible to the reader that large groups of people might be mistaken about their visual experience of perspective, and vice versa.  Likewise across kinds of experience: To the extent I can render it plausible that people are often badly mistaken about their visual imagery, it becomes less far-fetched to suppose they might be similarly radically mistaken about their emotional phenomenology, and vice versa.  Even if no single case is compelling on its own – and perhaps no case is entirely compelling on its own, given the variety of countermoves available to those who aim to defend the accuracy of introspection – I hope that the body of examples and arguments, considered as a whole, is still sufficient to make a good case for widespread introspective ignorance and error, even about what might have seemed on the surface to be easy and obvious matters.

I aim toward both a personal conclusion and a typical-person conclusion.  The personal conclusion is this: You the reader do not know your own stream of experience very well.  The doubt induction arguments – maybe they are better called “exercises” – are the main path to this conclusion.  However, I hope that the other two argument forms, when combined with the recognition of potentially mistaken others as your epistemic peers, can also lend considerable support to the personal conclusion.  In a similar spirit, I sometimes supplement my three main argument forms with confessional epistemology – sincere expression of my own degree of confidence or doubt on the matters at hand, offering myself as an epistemic peer.  I’ve found that most people default toward a kind of cavalier confidence when casually queried about their experience.  I aim to undercut that confidence, if you tend to feel it, and launch us together into the murks of doubt.

The typical-person conclusion is the conclusion that the typical person, or at least the typical person in our society, doesn’t know her own stream of experience very well.  The arguments from variability and error are the primary arguments toward this conclusion, though doubt induction can also lend support to this conclusion if the reader believes that she is in no worse an epistemic condition regarding her stream of experience than is the typical person.  The typical-person conclusion leaves open the possibility that some exceptional people have excellent knowledge of their stream of experience, through native talent or introspective training.

We are subject to a certain type of epistemic illusion.  Introspection seems easy, and we tend to have high confidence in its results, but that confidence is ill-tuned to epistemic reality.  Our intuitive sense of confidence, or even indubitability, exceeds what is warranted.  Some arguments for excellent self-knowledge – in Descartes (1641/1984), for example, and in Brie Gertler (2012) – appear to rely on these inflated epistemic intuitions, on an excessively self-trusting sense of the impossibility or near-impossibility of error.  We are invited to ask ourselves, how can I doubt this?  And from the fact that doubt seems impossible or possible only inauthetically, the epistemic authority of introspection follows.  But the arguments from variability and error should lead us to mistrust these feelings of confidence, if we see what seems to be misplaced confidence in our epistemic peers.  In fact, appeals to epistemic intuition can cut both ways: I use them toward skeptical ends in my doubt induction exercises.

  In principle, epistemic illusions could err either direction.  We could trend either too low or too high, either too underconfident or too overconfident.  But in fact, overconfidence is the more likely direction of error.  My arguments from variability and error suggest that people are often mistaken about their stream of conscious experience despite exhibiting outward signs of confidence.  Also, environmental pressures more often punish underconfidence (e.g., via looking inept or inarticulate or uninteresting when asked to discuss one’s visual imagery) than they punish overconfidence, since it is unusual to find naturally occurring situations in which people’s confident self-reports about their stream of experience are proven embarrassingly wrong.  In any case, meta-level doubts will generally tend to favor the skeptic in the end: If it is sufficiently unclear to you whether you should doubt P or not doubt P, you are already doubting P.[1]



Descartes, René (1641/1984).  Meditations on first philosophy.  In J. Cottingham, R. Stoothoff, and D. Murdoch, eds., The philosophical writings of Descartes, vol. II.  Cambridge: Cambridge.

Gertler, Brie (2012).  Renewed acquaintance.  In D. Smithies and D. Stoljar, eds., Introspection and consciousness.  Oxford: Oxford.

Schwitzgebel, Eric (2011).  Perplexities of consciousness.  Cambridge, MA: MIT.


[1] Thanks to Jakob Hohwy for helping me see my general argument forms, and to Jenny Windt for helping me see the central role of epistemic intuitions in my arguments.